diff --git a/AAAI/aaai24.bib b/AAAI/aaai24.bib
new file mode 100644
index 0000000000000000000000000000000000000000..7b7d2bcf44a7488d282c3e9b1f9079598dc99fa3
--- /dev/null
+++ b/AAAI/aaai24.bib
@@ -0,0 +1,111 @@
+@book{em:86,
+ editor = "Engelmore, Robert and Morgan, Anthony",
+ title = "Blackboard Systems",
+ year = 1986,
+ address = "Reading, Mass.",
+ publisher = "Addison-Wesley",
+}
+
+@inproceedings{c:83,
+ author = "Clancey, William J.",
+ year = 1983,
+ title = "{Communication, Simulation, and Intelligent
+Agents: Implications of Personal Intelligent Machines
+for Medical Education}",
+ booktitle="Proceedings of the Eighth International Joint Conference on Artificial Intelligence {(IJCAI-83)}",
+ pages = "556-560",
+ address = "Menlo Park, Calif",
+ publisher = "{IJCAI Organization}",
+}
+@inproceedings{c:84,
+ author = "Clancey, William J.",
+ year = 1984,
+ title = "{Classification Problem Solving}",
+ booktitle = "Proceedings of the Fourth National
+ Conference on Artificial Intelligence",
+ pages = "45-54",
+ address = "Menlo Park, Calif.",
+ publisher="AAAI Press",
+}
+@article{r:80,
+ author = {Robinson, Arthur L.},
+ title = {New Ways to Make Microcircuits Smaller},
+ volume = {208},
+ number = {4447},
+ pages = {1019--1022},
+ year = {1980},
+ doi = {10.1126/science.208.4447.1019},
+ publisher = {American Association for the Advancement of Science},
+ issn = {0036-8075},
+ URL = {https://science.sciencemag.org/content/208/4447/1019},
+ eprint = {https://science.sciencemag.org/content/208/4447/1019.full.pdf},
+ journal = {Science},
+}
+@article{r:80x,
+ author = "Robinson, Arthur L.",
+ year = 1980,
+ title = "{New Ways to Make Microcircuits Smaller---Duplicate Entry}",
+ journal = "Science",
+ volume = 208,
+ pages = "1019-1026",
+}
+@article{hcr:83,
+title = {Strategic explanations for a diagnostic consultation system},
+journal = {International Journal of Man-Machine Studies},
+volume = {20},
+number = {1},
+pages = {3-19},
+year = {1984},
+issn = {0020-7373},
+doi = {https://doi.org/10.1016/S0020-7373(84)80003-6},
+url = {https://www.sciencedirect.com/science/article/pii/S0020737384800036},
+author = {Diane Warner Hasling and William J. Clancey and Glenn Rennels},
+abstract = {This article examines the problem of automatte explanation of reasoning, especially as it relates to expert systems. By explanation we mean the ability of a program to discuss what it is doing in some understandable way. We first present a general framework in which to view explanation and review some of the research done in this area. We then focus on the explanation system for NEOMYCIN, a medical consultation program. A consultation program interactively helps a user to solve a problem. Our goal is to have NEOMYCIN explain its problem-solving strategies. An explanation of strategy describes the plan the program is using to reach a solution. Such an explanation is usually concrete, referring to aspects of the current problem situation. Abstract explanations articulate a general principle, which can be applied in different situations; such explanations are useful in teaching and in explaining by analogy. We describe the aspects of NEOMYCIN that make abstract strategic explanations possible—the representation of strategic knowledge explicitly and separately from domain knowledge— and demonstrate how this representation can be used to generate explanations.}
+}
+@article{hcrt:83,
+ author = "Hasling, Diane Warner and Clancey, William J. and Rennels, Glenn R. and Test, Thomas",
+ year = 1983,
+ title = "{Strategic Explanations in Consultation---Duplicate}",
+ journal = "The International Journal of Man-Machine Studies",
+ volume = 20,
+ number = 1,
+ pages = "3-19",
+}
+@techreport{r:86,
+ author = "Rice, James",
+ year = 1986,
+ title = "{Poligon: A System for Parallel Problem Solving}",
+ type = "Technical Report",
+ number = "KSL-86-19",
+ institution = "Dept.\ of Computer Science, Stanford Univ.",
+}
+@phdthesis{c:79,
+ author = "Clancey, William J.",
+ year = 1979,
+ title = "{Transfer of Rule-Based Expertise
+through a Tutorial Dialogue}",
+ type = "{Ph.D.} diss.",
+ school = "Dept.\ of Computer Science, Stanford Univ.",
+ address = "Stanford, Calif.",
+}
+@unpublished{c:21,
+ author = "Clancey, William J.",
+ title = "{The Engineering of Qualitative Models}",
+ year = 2021,
+ note = "Forthcoming",
+}
+@misc{c:22,
+ title={Attention Is All You Need},
+ author={Ashish Vaswani and Noam Shazeer and Niki Parmar and Jakob Uszkoreit and Llion Jones and Aidan N. Gomez and Lukasz Kaiser and Illia Polosukhin},
+ year={2017},
+ eprint={1706.03762},
+ archivePrefix={arXiv},
+ primaryClass={cs.CL}
+}
+@misc{c:23,
+ title = "Pluto: The 'Other' Red Planet",
+ author = "{NASA}",
+ howpublished = "\url{https://www.nasa.gov/nh/pluto-the-other-red-planet}",
+ year = 2015,
+ note = "Accessed: 2018-12-06"
+}
\ No newline at end of file
diff --git a/AAAI/aaai24.bst b/AAAI/aaai24.bst
new file mode 100644
index 0000000000000000000000000000000000000000..05b1d4e4414648ff46e9199f03bc66c9f77dedb5
--- /dev/null
+++ b/AAAI/aaai24.bst
@@ -0,0 +1,1493 @@
+%%
+%% This is file `aaai22.bst',
+%% generated with the docstrip utility.
+%%
+%% The original source files were:
+%%
+%% merlin.mbs (with options: `head,ay,nat,ed-au,nm-rev,ed-rev,jnrlst,aunm-semi,mcite,mct-1,mct-x3,keyxyr,dt-beg,yr-per,yrp-per,note-yr,atit-u,volp-sp,num-xser,bkpg-x,add-pub,isbn,ppx,ed,xedn,and-com,and-com-ed,etal-xc,nfss,,{}')
+%% merlin.mbs (with options: `tail,ay,nat,ed-au,nm-rev,ed-rev,jnrlst,aunm-semi,mcite,mct-1,mct-x3,keyxyr,dt-beg,yr-per,yrp-per,note-yr,atit-u,volp-sp,num-xser,bkpg-x,add-pub,isbn,ppx,ed,xedn,and-com,and-com-ed,etal-xc,nfss,,{}')
+%% ----------------------------------------
+%% *** Natbib-compatible implementation of 'aaai' bib style ***
+%%
+ % ===============================================================
+ % IMPORTANT NOTICE:
+ % This bibliographic style (bst) file has been generated from one or
+ % more master bibliographic style (mbs) files, listed above.
+ %
+ % This generated file can be redistributed and/or modified under the terms
+ % of the LaTeX Project Public License Distributed from CTAN
+ % archives in directory macros/latex/base/lppl.txt; either
+ % version 1 of the License, or any later version.
+ % ===============================================================
+ % Name and version information of the main mbs file:
+ % \ProvidesFile{merlin.mbs}[2011/11/18 4.33 (PWD, AO, DPC)]
+ % For use with BibTeX version 0.99a or later
+ %-------------------------------------------------------------------
+ % This bibliography style file is intended for texts in ENGLISH
+ % This is an author-year citation style bibliography. As such, it is
+ % non-standard LaTeX, and requires a special package file to function properly.
+ % Such a package is natbib.sty by Patrick W. Daly
+ % The form of the \bibitem entries is
+ % \bibitem[Jones et al.(1990)]{key}...
+ % \bibitem[Jones et al.(1990)Jones, Baker, and Smith]{key}...
+ % The essential feature is that the label (the part in brackets) consists
+ % of the author names, as they should appear in the citation, with the year
+ % in parentheses following. There must be no space before the opening
+ % parenthesis!
+ % With natbib v5.3, a full list of authors may also follow the year.
+ % In natbib.sty, it is possible to define the type of enclosures that is
+ % really wanted (brackets or parentheses), but in either case, there must
+ % be parentheses in the label.
+ % The \cite command functions as follows:
+ % \citet{key} ==>> Jones et al. (1990)
+ % \citet*{key} ==>> Jones, Baker, and Smith (1990)
+ % \citep{key} ==>> (Jones et al., 1990)
+ % \citep*{key} ==>> (Jones, Baker, and Smith, 1990)
+ % \citep[chap. 2]{key} ==>> (Jones et al., 1990, chap. 2)
+ % \citep[e.g.][]{key} ==>> (e.g. Jones et al., 1990)
+ % \citep[e.g.][p. 32]{key} ==>> (e.g. Jones et al., 1990, p. 32)
+ % \citeauthor{key} ==>> Jones et al.
+ % \citeauthor*{key} ==>> Jones, Baker, and Smith
+ % \citeyear{key} ==>> 1990
+ %---------------------------------------------------------------------
+
+ENTRY
+ { address
+ archivePrefix
+ author
+ booktitle
+ chapter
+ edition
+ editor
+ eid
+ eprint
+ howpublished
+ institution
+ isbn
+ journal
+ key
+ month
+ note
+ number
+ organization
+ pages
+ publisher
+ school
+ series
+ title
+ type
+ volume
+ year
+ }
+ {}
+ { label extra.label sort.label short.list }
+INTEGERS { output.state before.all mid.sentence after.sentence after.block }
+FUNCTION {init.state.consts}
+{ #0 'before.all :=
+ #1 'mid.sentence :=
+ #2 'after.sentence :=
+ #3 'after.block :=
+}
+STRINGS { s t}
+FUNCTION {output.nonnull}
+{ 's :=
+ output.state mid.sentence =
+ { ", " * write$ }
+ { output.state after.block =
+ { add.period$ write$
+ newline$
+ "\newblock " write$
+ }
+ { output.state before.all =
+ 'write$
+ { add.period$ " " * write$ }
+ if$
+ }
+ if$
+ mid.sentence 'output.state :=
+ }
+ if$
+ s
+}
+FUNCTION {output}
+{ duplicate$ empty$
+ 'pop$
+ 'output.nonnull
+ if$
+}
+FUNCTION {output.check}
+{ 't :=
+ duplicate$ empty$
+ { pop$ "empty " t * " in " * cite$ * warning$ }
+ 'output.nonnull
+ if$
+}
+FUNCTION {fin.entry}
+{ add.period$
+ write$
+ newline$
+}
+
+FUNCTION {new.block}
+{ output.state before.all =
+ 'skip$
+ { after.block 'output.state := }
+ if$
+}
+FUNCTION {new.sentence}
+{ output.state after.block =
+ 'skip$
+ { output.state before.all =
+ 'skip$
+ { after.sentence 'output.state := }
+ if$
+ }
+ if$
+}
+FUNCTION {add.blank}
+{ " " * before.all 'output.state :=
+}
+
+FUNCTION {date.block}
+{
+ new.block
+}
+
+FUNCTION {not}
+{ { #0 }
+ { #1 }
+ if$
+}
+FUNCTION {and}
+{ 'skip$
+ { pop$ #0 }
+ if$
+}
+FUNCTION {or}
+{ { pop$ #1 }
+ 'skip$
+ if$
+}
+FUNCTION {new.block.checkb}
+{ empty$
+ swap$ empty$
+ and
+ 'skip$
+ 'new.block
+ if$
+}
+FUNCTION {field.or.null}
+{ duplicate$ empty$
+ { pop$ "" }
+ 'skip$
+ if$
+}
+FUNCTION {emphasize}
+{ duplicate$ empty$
+ { pop$ "" }
+ { "\emph{" swap$ * "}" * }
+ if$
+}
+FUNCTION {tie.or.space.prefix}
+{ duplicate$ text.length$ #3 <
+ { "~" }
+ { " " }
+ if$
+ swap$
+}
+
+FUNCTION {capitalize}
+{ "u" change.case$ "t" change.case$ }
+
+FUNCTION {space.word}
+{ " " swap$ * " " * }
+ % Here are the language-specific definitions for explicit words.
+ % Each function has a name bbl.xxx where xxx is the English word.
+ % The language selected here is ENGLISH
+FUNCTION {bbl.and}
+{ "and"}
+
+FUNCTION {bbl.etal}
+{ "et~al." }
+
+FUNCTION {bbl.editors}
+{ "eds." }
+
+FUNCTION {bbl.editor}
+{ "ed." }
+
+FUNCTION {bbl.edby}
+{ "edited by" }
+
+FUNCTION {bbl.edition}
+{ "edition" }
+
+FUNCTION {bbl.volume}
+{ "volume" }
+
+FUNCTION {bbl.of}
+{ "of" }
+
+FUNCTION {bbl.number}
+{ "number" }
+
+FUNCTION {bbl.nr}
+{ "no." }
+
+FUNCTION {bbl.in}
+{ "in" }
+
+FUNCTION {bbl.pages}
+{ "" }
+
+FUNCTION {bbl.page}
+{ "" }
+
+FUNCTION {bbl.chapter}
+{ "chapter" }
+
+FUNCTION {bbl.techrep}
+{ "Technical Report" }
+
+FUNCTION {bbl.mthesis}
+{ "Master's thesis" }
+
+FUNCTION {bbl.phdthesis}
+{ "Ph.D. thesis" }
+
+MACRO {jan} {"January"}
+
+MACRO {feb} {"February"}
+
+MACRO {mar} {"March"}
+
+MACRO {apr} {"April"}
+
+MACRO {may} {"May"}
+
+MACRO {jun} {"June"}
+
+MACRO {jul} {"July"}
+
+MACRO {aug} {"August"}
+
+MACRO {sep} {"September"}
+
+MACRO {oct} {"October"}
+
+MACRO {nov} {"November"}
+
+MACRO {dec} {"December"}
+
+MACRO {acmcs} {"ACM Computing Surveys"}
+
+MACRO {acta} {"Acta Informatica"}
+
+MACRO {cacm} {"Communications of the ACM"}
+
+MACRO {ibmjrd} {"IBM Journal of Research and Development"}
+
+MACRO {ibmsj} {"IBM Systems Journal"}
+
+MACRO {ieeese} {"IEEE Transactions on Software Engineering"}
+
+MACRO {ieeetc} {"IEEE Transactions on Computers"}
+
+MACRO {ieeetcad}
+ {"IEEE Transactions on Computer-Aided Design of Integrated Circuits"}
+
+MACRO {ipl} {"Information Processing Letters"}
+
+MACRO {jacm} {"Journal of the ACM"}
+
+MACRO {jcss} {"Journal of Computer and System Sciences"}
+
+MACRO {scp} {"Science of Computer Programming"}
+
+MACRO {sicomp} {"SIAM Journal on Computing"}
+
+MACRO {tocs} {"ACM Transactions on Computer Systems"}
+
+MACRO {tods} {"ACM Transactions on Database Systems"}
+
+MACRO {tog} {"ACM Transactions on Graphics"}
+
+MACRO {toms} {"ACM Transactions on Mathematical Software"}
+
+MACRO {toois} {"ACM Transactions on Office Information Systems"}
+
+MACRO {toplas} {"ACM Transactions on Programming Languages and Systems"}
+
+MACRO {tcs} {"Theoretical Computer Science"}
+FUNCTION {bibinfo.check}
+{ swap$
+ duplicate$ missing$
+ {
+ pop$ pop$
+ ""
+ }
+ { duplicate$ empty$
+ {
+ swap$ pop$
+ }
+ { swap$
+ pop$
+ }
+ if$
+ }
+ if$
+}
+FUNCTION {bibinfo.warn}
+{ swap$
+ duplicate$ missing$
+ {
+ swap$ "missing " swap$ * " in " * cite$ * warning$ pop$
+ ""
+ }
+ { duplicate$ empty$
+ {
+ swap$ "empty " swap$ * " in " * cite$ * warning$
+ }
+ { swap$
+ pop$
+ }
+ if$
+ }
+ if$
+}
+FUNCTION {format.eprint}
+{ eprint duplicate$ empty$
+ 'skip$
+ { archivePrefix duplicate$ empty$
+ 'skip$
+ { ":" * swap$ }
+ if$
+ * "." *
+ }
+ if$
+}
+INTEGERS { nameptr namesleft numnames }
+
+
+STRINGS { bibinfo}
+
+FUNCTION {format.names}
+{ 'bibinfo :=
+ duplicate$ empty$ 'skip$ {
+ 's :=
+ "" 't :=
+ #1 'nameptr :=
+ s num.names$ 'numnames :=
+ numnames 'namesleft :=
+ { namesleft #0 > }
+ { s nameptr
+ "{vv~}{ll}{, f.}{, jj}"
+ format.name$
+ bibinfo bibinfo.check
+ 't :=
+ nameptr #1 >
+ {
+ namesleft #1 >
+ { "; " * t * }
+ {
+ s nameptr "{ll}" format.name$ duplicate$ "others" =
+ { 't := }
+ { pop$ }
+ if$
+ ";" *
+ t "others" =
+ {
+ " " * bbl.etal *
+ }
+ {
+ bbl.and
+ space.word * t *
+ }
+ if$
+ }
+ if$
+ }
+ 't
+ if$
+ nameptr #1 + 'nameptr :=
+ namesleft #1 - 'namesleft :=
+ }
+ while$
+ } if$
+}
+FUNCTION {format.names.ed}
+{
+ format.names
+}
+FUNCTION {format.key}
+{ empty$
+ { key field.or.null }
+ { "" }
+ if$
+}
+
+FUNCTION {format.authors}
+{ author "author" format.names
+}
+FUNCTION {get.bbl.editor}
+{ editor num.names$ #1 > 'bbl.editors 'bbl.editor if$ }
+
+FUNCTION {format.editors}
+{ editor "editor" format.names duplicate$ empty$ 'skip$
+ {
+ "," *
+ " " *
+ get.bbl.editor
+ *
+ }
+ if$
+}
+FUNCTION {format.isbn}
+{ isbn "isbn" bibinfo.check
+ duplicate$ empty$ 'skip$
+ {
+ new.block
+ "ISBN " swap$ *
+ }
+ if$
+}
+
+FUNCTION {format.note}
+{
+ note empty$
+ { "" }
+ { note #1 #1 substring$
+ duplicate$ "{" =
+ 'skip$
+ { output.state mid.sentence =
+ { "l" }
+ { "u" }
+ if$
+ change.case$
+ }
+ if$
+ note #2 global.max$ substring$ * "note" bibinfo.check
+ }
+ if$
+}
+
+FUNCTION {format.title}
+{ title
+ "title" bibinfo.check
+}
+FUNCTION {format.full.names}
+{'s :=
+ "" 't :=
+ #1 'nameptr :=
+ s num.names$ 'numnames :=
+ numnames 'namesleft :=
+ { namesleft #0 > }
+ { s nameptr
+ "{vv~}{ll}" format.name$
+ 't :=
+ nameptr #1 >
+ {
+ namesleft #1 >
+ { ", " * t * }
+ {
+ s nameptr "{ll}" format.name$ duplicate$ "others" =
+ { 't := }
+ { pop$ }
+ if$
+ t "others" =
+ {
+ " " * bbl.etal *
+ }
+ {
+ numnames #2 >
+ { "," * }
+ 'skip$
+ if$
+ bbl.and
+ space.word * t *
+ }
+ if$
+ }
+ if$
+ }
+ 't
+ if$
+ nameptr #1 + 'nameptr :=
+ namesleft #1 - 'namesleft :=
+ }
+ while$
+}
+
+FUNCTION {author.editor.key.full}
+{ author empty$
+ { editor empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { editor format.full.names }
+ if$
+ }
+ { author format.full.names }
+ if$
+}
+
+FUNCTION {author.key.full}
+{ author empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { author format.full.names }
+ if$
+}
+
+FUNCTION {editor.key.full}
+{ editor empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { editor format.full.names }
+ if$
+}
+
+FUNCTION {make.full.names}
+{ type$ "book" =
+ type$ "inbook" =
+ or
+ 'author.editor.key.full
+ { type$ "proceedings" =
+ 'editor.key.full
+ 'author.key.full
+ if$
+ }
+ if$
+}
+
+FUNCTION {output.bibitem}
+{ newline$
+ "\bibitem[{" write$
+ label write$
+ ")" make.full.names duplicate$ short.list =
+ { pop$ }
+ { * }
+ if$
+ "}]{" * write$
+ cite$ write$
+ "}" write$
+ newline$
+ ""
+ before.all 'output.state :=
+}
+
+FUNCTION {n.dashify}
+{
+ 't :=
+ ""
+ { t empty$ not }
+ { t #1 #1 substring$ "-" =
+ { t #1 #2 substring$ "--" = not
+ { "--" *
+ t #2 global.max$ substring$ 't :=
+ }
+ { { t #1 #1 substring$ "-" = }
+ { "-" *
+ t #2 global.max$ substring$ 't :=
+ }
+ while$
+ }
+ if$
+ }
+ { t #1 #1 substring$ *
+ t #2 global.max$ substring$ 't :=
+ }
+ if$
+ }
+ while$
+}
+
+FUNCTION {word.in}
+{ bbl.in capitalize
+ " " * }
+
+FUNCTION {format.date}
+{ year "year" bibinfo.check duplicate$ empty$
+ {
+ "empty year in " cite$ * "; set to ????" * warning$
+ pop$ "????"
+ }
+ 'skip$
+ if$
+ extra.label *
+ before.all 'output.state :=
+ after.sentence 'output.state :=
+}
+FUNCTION {format.btitle}
+{ title "title" bibinfo.check
+ duplicate$ empty$ 'skip$
+ {
+ emphasize
+ }
+ if$
+}
+FUNCTION {either.or.check}
+{ empty$
+ 'pop$
+ { "can't use both " swap$ * " fields in " * cite$ * warning$ }
+ if$
+}
+FUNCTION {format.bvolume}
+{ volume empty$
+ { "" }
+ { bbl.volume volume tie.or.space.prefix
+ "volume" bibinfo.check * *
+ series "series" bibinfo.check
+ duplicate$ empty$ 'pop$
+ { swap$ bbl.of space.word * swap$
+ emphasize * }
+ if$
+ "volume and number" number either.or.check
+ }
+ if$
+}
+FUNCTION {format.number.series}
+{ volume empty$
+ { number empty$
+ { series field.or.null }
+ { series empty$
+ { number "number" bibinfo.check }
+ { output.state mid.sentence =
+ { bbl.number }
+ { bbl.number capitalize }
+ if$
+ number tie.or.space.prefix "number" bibinfo.check * *
+ bbl.in space.word *
+ series "series" bibinfo.check *
+ }
+ if$
+ }
+ if$
+ }
+ { "" }
+ if$
+}
+
+FUNCTION {format.edition}
+{ edition duplicate$ empty$ 'skip$
+ {
+ output.state mid.sentence =
+ { "l" }
+ { "t" }
+ if$ change.case$
+ "edition" bibinfo.check
+ " " * bbl.edition *
+ }
+ if$
+}
+INTEGERS { multiresult }
+FUNCTION {multi.page.check}
+{ 't :=
+ #0 'multiresult :=
+ { multiresult not
+ t empty$ not
+ and
+ }
+ { t #1 #1 substring$
+ duplicate$ "-" =
+ swap$ duplicate$ "," =
+ swap$ "+" =
+ or or
+ { #1 'multiresult := }
+ { t #2 global.max$ substring$ 't := }
+ if$
+ }
+ while$
+ multiresult
+}
+FUNCTION {format.pages}
+{ pages duplicate$ empty$ 'skip$
+ { duplicate$ multi.page.check
+ {
+ n.dashify
+ }
+ {
+ }
+ if$
+ "pages" bibinfo.check
+ }
+ if$
+}
+FUNCTION {format.journal.pages}
+{ pages duplicate$ empty$ 'pop$
+ { swap$ duplicate$ empty$
+ { pop$ pop$ format.pages }
+ {
+ ": " *
+ swap$
+ n.dashify
+ "pages" bibinfo.check
+ *
+ }
+ if$
+ }
+ if$
+}
+FUNCTION {format.journal.eid}
+{ eid "eid" bibinfo.check
+ duplicate$ empty$ 'pop$
+ { swap$ duplicate$ empty$ 'skip$
+ {
+ ": " *
+ }
+ if$
+ swap$ *
+ }
+ if$
+}
+FUNCTION {format.vol.num.pages}
+{ volume field.or.null
+ duplicate$ empty$ 'skip$
+ {
+ "volume" bibinfo.check
+ }
+ if$
+ number "number" bibinfo.check duplicate$ empty$ 'skip$
+ {
+ swap$ duplicate$ empty$
+ { "there's a number but no volume in " cite$ * warning$ }
+ 'skip$
+ if$
+ swap$
+ "(" swap$ * ")" *
+ }
+ if$ *
+ eid empty$
+ { format.journal.pages }
+ { format.journal.eid }
+ if$
+}
+
+FUNCTION {format.chapter.pages}
+{ chapter empty$
+ 'format.pages
+ { type empty$
+ { bbl.chapter }
+ { type "l" change.case$
+ "type" bibinfo.check
+ }
+ if$
+ chapter tie.or.space.prefix
+ "chapter" bibinfo.check
+ * *
+ pages empty$
+ 'skip$
+ { ", " * format.pages * }
+ if$
+ }
+ if$
+}
+
+FUNCTION {format.booktitle}
+{
+ booktitle "booktitle" bibinfo.check
+ emphasize
+}
+FUNCTION {format.in.ed.booktitle}
+{ format.booktitle duplicate$ empty$ 'skip$
+ {
+ editor "editor" format.names.ed duplicate$ empty$ 'pop$
+ {
+ "," *
+ " " *
+ get.bbl.editor
+ ", " *
+ * swap$
+ * }
+ if$
+ word.in swap$ *
+ }
+ if$
+}
+FUNCTION {format.thesis.type}
+{ type duplicate$ empty$
+ 'pop$
+ { swap$ pop$
+ "t" change.case$ "type" bibinfo.check
+ }
+ if$
+}
+FUNCTION {format.tr.number}
+{ number "number" bibinfo.check
+ type duplicate$ empty$
+ { pop$ bbl.techrep }
+ 'skip$
+ if$
+ "type" bibinfo.check
+ swap$ duplicate$ empty$
+ { pop$ "t" change.case$ }
+ { tie.or.space.prefix * * }
+ if$
+}
+FUNCTION {format.article.crossref}
+{
+ word.in
+ " \cite{" * crossref * "}" *
+}
+FUNCTION {format.book.crossref}
+{ volume duplicate$ empty$
+ { "empty volume in " cite$ * "'s crossref of " * crossref * warning$
+ pop$ word.in
+ }
+ { bbl.volume
+ capitalize
+ swap$ tie.or.space.prefix "volume" bibinfo.check * * bbl.of space.word *
+ }
+ if$
+ " \cite{" * crossref * "}" *
+}
+FUNCTION {format.incoll.inproc.crossref}
+{
+ word.in
+ " \cite{" * crossref * "}" *
+}
+FUNCTION {format.org.or.pub}
+{ 't :=
+ ""
+ address empty$ t empty$ and
+ 'skip$
+ {
+ address "address" bibinfo.check *
+ t empty$
+ 'skip$
+ { address empty$
+ 'skip$
+ { ": " * }
+ if$
+ t *
+ }
+ if$
+ }
+ if$
+}
+FUNCTION {format.publisher.address}
+{ publisher "publisher" bibinfo.warn format.org.or.pub
+}
+
+FUNCTION {format.organization.address}
+{ organization "organization" bibinfo.check format.org.or.pub
+}
+
+FUNCTION {article}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title "title" output.check
+ new.block
+ crossref missing$
+ {
+ journal
+ "journal" bibinfo.check
+ emphasize
+ "journal" output.check
+ format.vol.num.pages output
+ }
+ { format.article.crossref output.nonnull
+ format.pages output
+ }
+ if$
+ new.block
+ format.note output
+ fin.entry
+}
+FUNCTION {book}
+{ output.bibitem
+ author empty$
+ { format.editors "author and editor" output.check
+ editor format.key output
+ }
+ { format.authors output.nonnull
+ crossref missing$
+ { "author and editor" editor either.or.check }
+ 'skip$
+ if$
+ }
+ if$
+ format.date "year" output.check
+ date.block
+ format.btitle "title" output.check
+ crossref missing$
+ { format.bvolume output
+ new.block
+ format.number.series output
+ new.sentence
+ format.publisher.address output
+ }
+ {
+ new.block
+ format.book.crossref output.nonnull
+ }
+ if$
+ format.edition output
+ format.isbn output
+ new.block
+ format.note output
+ fin.entry
+}
+FUNCTION {booklet}
+{ output.bibitem
+ format.authors output
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title "title" output.check
+ new.block
+ howpublished "howpublished" bibinfo.check output
+ address "address" bibinfo.check output
+ format.isbn output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {inbook}
+{ output.bibitem
+ author empty$
+ { format.editors "author and editor" output.check
+ editor format.key output
+ }
+ { format.authors output.nonnull
+ crossref missing$
+ { "author and editor" editor either.or.check }
+ 'skip$
+ if$
+ }
+ if$
+ format.date "year" output.check
+ date.block
+ format.btitle "title" output.check
+ crossref missing$
+ {
+ format.bvolume output
+ format.chapter.pages "chapter and pages" output.check
+ new.block
+ format.number.series output
+ new.sentence
+ format.publisher.address output
+ }
+ {
+ format.chapter.pages "chapter and pages" output.check
+ new.block
+ format.book.crossref output.nonnull
+ }
+ if$
+ format.edition output
+ crossref missing$
+ { format.isbn output }
+ 'skip$
+ if$
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {incollection}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title "title" output.check
+ new.block
+ crossref missing$
+ { format.in.ed.booktitle "booktitle" output.check
+ format.bvolume output
+ format.number.series output
+ format.chapter.pages output
+ new.sentence
+ format.publisher.address output
+ format.edition output
+ format.isbn output
+ }
+ { format.incoll.inproc.crossref output.nonnull
+ format.chapter.pages output
+ }
+ if$
+ new.block
+ format.note output
+ fin.entry
+}
+FUNCTION {inproceedings}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title "title" output.check
+ new.block
+ crossref missing$
+ { format.in.ed.booktitle "booktitle" output.check
+ format.bvolume output
+ format.number.series output
+ format.pages output
+ new.sentence
+ publisher empty$
+ { format.organization.address output }
+ { organization "organization" bibinfo.check output
+ format.publisher.address output
+ }
+ if$
+ format.isbn output
+ }
+ { format.incoll.inproc.crossref output.nonnull
+ format.pages output
+ }
+ if$
+ new.block
+ format.note output
+ fin.entry
+}
+FUNCTION {conference} { inproceedings }
+FUNCTION {manual}
+{ output.bibitem
+ format.authors output
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.btitle "title" output.check
+ organization address new.block.checkb
+ organization "organization" bibinfo.check output
+ address "address" bibinfo.check output
+ format.edition output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {mastersthesis}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.btitle
+ "title" output.check
+ new.block
+ bbl.mthesis format.thesis.type output.nonnull
+ school "school" bibinfo.warn output
+ address "address" bibinfo.check output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {misc}
+{ output.bibitem
+ format.authors output
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title output
+ new.block
+ howpublished "howpublished" bibinfo.check output
+ new.block
+ format.note output
+ format.eprint output
+ fin.entry
+}
+FUNCTION {phdthesis}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.btitle
+ "title" output.check
+ new.block
+ bbl.phdthesis format.thesis.type output.nonnull
+ school "school" bibinfo.warn output
+ address "address" bibinfo.check output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {proceedings}
+{ output.bibitem
+ format.editors output
+ editor format.key output
+ format.date "year" output.check
+ date.block
+ format.btitle "title" output.check
+ format.bvolume output
+ format.number.series output
+ new.sentence
+ publisher empty$
+ { format.organization.address output }
+ { organization "organization" bibinfo.check output
+ format.publisher.address output
+ }
+ if$
+ format.isbn output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {techreport}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title
+ "title" output.check
+ new.block
+ format.tr.number output.nonnull
+ institution "institution" bibinfo.warn output
+ address "address" bibinfo.check output
+ new.block
+ format.note output
+ fin.entry
+}
+
+FUNCTION {unpublished}
+{ output.bibitem
+ format.authors "author" output.check
+ author format.key output
+ format.date "year" output.check
+ date.block
+ format.title "title" output.check
+ new.block
+ format.note "note" output.check
+ fin.entry
+}
+
+FUNCTION {default.type} { misc }
+READ
+FUNCTION {sortify}
+{ purify$
+ "l" change.case$
+}
+INTEGERS { len }
+FUNCTION {chop.word}
+{ 's :=
+ 'len :=
+ s #1 len substring$ =
+ { s len #1 + global.max$ substring$ }
+ 's
+ if$
+}
+FUNCTION {format.lab.names}
+{'s :=
+ "" 't :=
+ #1 'nameptr :=
+ s num.names$ 'numnames :=
+ numnames 'namesleft :=
+ { namesleft #0 > }
+ { s nameptr
+ "{vv~}{ll}" format.name$
+ 't :=
+ nameptr #1 >
+ {
+ nameptr #2 =
+ numnames #3 > and
+ { "others" 't :=
+ #1 'namesleft := }
+ 'skip$
+ if$
+ namesleft #1 >
+ { ", " * t * }
+ {
+ s nameptr "{ll}" format.name$ duplicate$ "others" =
+ { 't := }
+ { pop$ }
+ if$
+ t "others" =
+ {
+ " " * bbl.etal *
+ }
+ {
+ numnames #2 >
+ { "," * }
+ 'skip$
+ if$
+ bbl.and
+ space.word * t *
+ }
+ if$
+ }
+ if$
+ }
+ 't
+ if$
+ nameptr #1 + 'nameptr :=
+ namesleft #1 - 'namesleft :=
+ }
+ while$
+}
+
+FUNCTION {author.key.label}
+{ author empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { author format.lab.names }
+ if$
+}
+
+FUNCTION {author.editor.key.label}
+{ author empty$
+ { editor empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { editor format.lab.names }
+ if$
+ }
+ { author format.lab.names }
+ if$
+}
+
+FUNCTION {editor.key.label}
+{ editor empty$
+ { key empty$
+ { cite$ #1 #3 substring$ }
+ 'key
+ if$
+ }
+ { editor format.lab.names }
+ if$
+}
+
+FUNCTION {calc.short.authors}
+{ type$ "book" =
+ type$ "inbook" =
+ or
+ 'author.editor.key.label
+ { type$ "proceedings" =
+ 'editor.key.label
+ 'author.key.label
+ if$
+ }
+ if$
+ 'short.list :=
+}
+
+FUNCTION {calc.label}
+{ calc.short.authors
+ short.list
+ "("
+ *
+ year duplicate$ empty$
+ short.list key field.or.null = or
+ { pop$ "" }
+ 'skip$
+ if$
+ *
+ 'label :=
+}
+
+FUNCTION {sort.format.names}
+{ 's :=
+ #1 'nameptr :=
+ ""
+ s num.names$ 'numnames :=
+ numnames 'namesleft :=
+ { namesleft #0 > }
+ { s nameptr
+ "{vv{ } }{ll{ }}{ f{ }}{ jj{ }}"
+ format.name$ 't :=
+ nameptr #1 >
+ {
+ " " *
+ namesleft #1 = t "others" = and
+ { "zzzzz" 't := }
+ 'skip$
+ if$
+ t sortify *
+ }
+ { t sortify * }
+ if$
+ nameptr #1 + 'nameptr :=
+ namesleft #1 - 'namesleft :=
+ }
+ while$
+}
+
+FUNCTION {sort.format.title}
+{ 't :=
+ "A " #2
+ "An " #3
+ "The " #4 t chop.word
+ chop.word
+ chop.word
+ sortify
+ #1 global.max$ substring$
+}
+FUNCTION {author.sort}
+{ author empty$
+ { key empty$
+ { "to sort, need author or key in " cite$ * warning$
+ ""
+ }
+ { key sortify }
+ if$
+ }
+ { author sort.format.names }
+ if$
+}
+FUNCTION {author.editor.sort}
+{ author empty$
+ { editor empty$
+ { key empty$
+ { "to sort, need author, editor, or key in " cite$ * warning$
+ ""
+ }
+ { key sortify }
+ if$
+ }
+ { editor sort.format.names }
+ if$
+ }
+ { author sort.format.names }
+ if$
+}
+FUNCTION {editor.sort}
+{ editor empty$
+ { key empty$
+ { "to sort, need editor or key in " cite$ * warning$
+ ""
+ }
+ { key sortify }
+ if$
+ }
+ { editor sort.format.names }
+ if$
+}
+FUNCTION {presort}
+{ calc.label
+ label sortify
+ " "
+ *
+ type$ "book" =
+ type$ "inbook" =
+ or
+ 'author.editor.sort
+ { type$ "proceedings" =
+ 'editor.sort
+ 'author.sort
+ if$
+ }
+ if$
+ #1 entry.max$ substring$
+ 'sort.label :=
+ sort.label
+ *
+ " "
+ *
+ title field.or.null
+ sort.format.title
+ *
+ #1 entry.max$ substring$
+ 'sort.key$ :=
+}
+
+ITERATE {presort}
+SORT
+STRINGS { last.label next.extra }
+INTEGERS { last.extra.num last.extra.num.extended last.extra.num.blank number.label }
+FUNCTION {initialize.extra.label.stuff}
+{ #0 int.to.chr$ 'last.label :=
+ "" 'next.extra :=
+ #0 'last.extra.num :=
+ "a" chr.to.int$ #1 - 'last.extra.num.blank :=
+ last.extra.num.blank 'last.extra.num.extended :=
+ #0 'number.label :=
+}
+FUNCTION {forward.pass}
+{ last.label label =
+ { last.extra.num #1 + 'last.extra.num :=
+ last.extra.num "z" chr.to.int$ >
+ { "a" chr.to.int$ 'last.extra.num :=
+ last.extra.num.extended #1 + 'last.extra.num.extended :=
+ }
+ 'skip$
+ if$
+ last.extra.num.extended last.extra.num.blank >
+ { last.extra.num.extended int.to.chr$
+ last.extra.num int.to.chr$
+ * 'extra.label := }
+ { last.extra.num int.to.chr$ 'extra.label := }
+ if$
+ }
+ { "a" chr.to.int$ 'last.extra.num :=
+ "" 'extra.label :=
+ label 'last.label :=
+ }
+ if$
+ number.label #1 + 'number.label :=
+}
+FUNCTION {reverse.pass}
+{ next.extra "b" =
+ { "a" 'extra.label := }
+ 'skip$
+ if$
+ extra.label 'next.extra :=
+ extra.label
+ duplicate$ empty$
+ 'skip$
+ { "{\natexlab{" swap$ * "}}" * }
+ if$
+ 'extra.label :=
+ label extra.label * 'label :=
+}
+EXECUTE {initialize.extra.label.stuff}
+ITERATE {forward.pass}
+REVERSE {reverse.pass}
+FUNCTION {bib.sort.order}
+{ sort.label
+ " "
+ *
+ year field.or.null sortify
+ *
+ " "
+ *
+ title field.or.null
+ sort.format.title
+ *
+ #1 entry.max$ substring$
+ 'sort.key$ :=
+}
+ITERATE {bib.sort.order}
+SORT
+FUNCTION {begin.bib}
+{ preamble$ empty$
+ 'skip$
+ { preamble$ write$ newline$ }
+ if$
+ "\begin{thebibliography}{" number.label int.to.str$ * "}" *
+ write$ newline$
+ "\providecommand{\natexlab}[1]{#1}"
+ write$ newline$
+}
+EXECUTE {begin.bib}
+EXECUTE {init.state.consts}
+ITERATE {call.type$}
+FUNCTION {end.bib}
+{ newline$
+ "\end{thebibliography}" write$ newline$
+}
+EXECUTE {end.bib}
+%% End of customized bst file
+%%
+%% End of file `aaai22.bst'.
diff --git a/AAAI/aaai24.sty b/AAAI/aaai24.sty
new file mode 100644
index 0000000000000000000000000000000000000000..a68f6036d6c04077479ab116f206e30e95e6479f
--- /dev/null
+++ b/AAAI/aaai24.sty
@@ -0,0 +1,303 @@
+\NeedsTeXFormat{LaTeX2e}%
+\ProvidesPackage{aaai24}[2023/06/26 AAAI 2024 Submission format]%
+\def\year{2024}%
+\typeout{Conference Style for AAAI for LaTeX 2e -- version for submission}%
+%
+\def\copyright@on{T}
+\def\showauthors@on{T}
+\def\nocopyright{\gdef\copyright@on{}} % Copyright notice is required for camera-ready only.
+\DeclareOption{submission}{%
+ \gdef\copyright@on{}%
+ \gdef\showauthors@on{}%
+ \long\gdef\pdfinfo #1{\relax}%
+}%
+\ProcessOptions\relax%
+% WARNING: IF YOU ARE USING THIS STYLE SHEET FOR AN AAAI PUBLICATION, YOU
+% MAY NOT MODIFY IT FOR ANY REASON. MODIFICATIONS (IN YOUR SOURCE
+% OR IN THIS STYLE SHEET WILL RESULT IN REJECTION OF YOUR PAPER).
+%
+% WARNING: This style is NOT guaranteed to work. It is provided in the
+% hope that it might make the preparation of papers easier, but this style
+% file is provided "as is" without warranty of any kind, either express or
+% implied, including but not limited to the implied warranties of
+% merchantability, fitness for a particular purpose, or noninfringement.
+% You use this style file at your own risk. Standard disclaimers apply.
+% There are undoubtably bugs in this style. If you would like to submit
+% bug fixes, improvements, etc. please let us know. Please use the contact form
+% at www.aaai.org.
+%
+% Do not use this file unless you are an experienced LaTeX user.
+%
+% PHYSICAL PAGE LAYOUT
+\setlength\topmargin{-0.25in} \setlength\oddsidemargin{-0.25in}
+\setlength\textheight{9.0in} \setlength\textwidth{7.0in}
+\setlength\columnsep{0.375in} \newlength\titlebox \setlength\titlebox{2.25in}
+\setlength\headheight{0pt} \setlength\headsep{0pt}
+%\setlength\footheight{0pt} \setlength\footskip{0pt}
+\thispagestyle{empty} \pagestyle{empty}
+\flushbottom \twocolumn \sloppy
+% We're never going to need a table of contents, so just flush it to
+% save space --- suggested by drstrip@sandia-2
+\def\addcontentsline#1#2#3{}
+% gf: PRINT COPYRIGHT NOTICE
+\def\copyright@year{\number\year}
+\def\copyright@text{Copyright \copyright\space \copyright@year,
+Association for the Advancement of Artificial Intelligence (www.aaai.org).
+All rights reserved.}
+\def\copyrighttext#1{\gdef\copyright@on{T}\gdef\copyright@text{#1}}
+\def\copyrightyear#1{\gdef\copyright@on{T}\gdef\copyright@year{#1}}
+% gf: End changes for copyright notice (used in \maketitle, below)
+% Title stuff, taken from deproc.
+%
+\def\maketitle{%
+ \par%
+ \begingroup % to make the footnote style local to the title
+ \def\thefootnote{\fnsymbol{footnote}}
+ \twocolumn[\@maketitle] \@thanks%
+ \endgroup%
+ % Insert copyright slug unless turned off
+ \if T\copyright@on\insert\footins{\noindent\footnotesize\copyright@text}\fi%
+ %
+ \setcounter{footnote}{0}%
+ \let\maketitle\relax%
+ \let\@maketitle\relax%
+ \gdef\@thanks{}%
+ \gdef\@author{}%
+ \gdef\@title{}%
+ \let\thanks\relax%
+}%
+\long\gdef\affiliations #1{ \def \affiliations_{\if T\showauthors@on#1\fi}}%
+%
+\def\@maketitle{%
+ \def\theauthors{\if T\showauthors@on\@author\else Anonymous submission\fi}
+ \newcounter{eqfn}\setcounter{eqfn}{0}%
+ \newsavebox{\titlearea}
+ \sbox{\titlearea}{
+ \let\footnote\relax\let\thanks\relax%
+ \setcounter{footnote}{0}%
+ \def\equalcontrib{%
+ \ifnum\value{eqfn}=0%
+ \footnote{These authors contributed equally.}%
+ \setcounter{eqfn}{\value{footnote}}%
+ \else%
+ \footnotemark[\value{eqfn}]%
+ \fi%
+ }%
+ \vbox{%
+ \hsize\textwidth%
+ \linewidth\hsize%
+ \vskip 0.625in minus 0.125in%
+ \centering%
+ {\LARGE\bf \@title \par}%
+ \vskip 0.1in plus 0.5fil minus 0.05in%
+ {\Large{\textbf{\theauthors\ifhmode\\\fi}}}%
+ \vskip .2em plus 0.25fil%
+ {\normalsize \affiliations_\ifhmode\\\fi}%
+ \vskip .5em plus 2fil%
+ }%
+ }%
+%
+ \newlength\actualheight%
+ \settoheight{\actualheight}{\usebox{\titlearea}}%
+ \ifdim\actualheight>\titlebox%
+ \setlength{\titlebox}{\actualheight}%
+ \fi%
+%
+ \vbox to \titlebox {%
+ \let\footnote\thanks\relax%
+ \setcounter{footnote}{0}%
+ \def\equalcontrib{%
+ \ifnum\value{eqfn}=0%
+ \footnote{These authors contributed equally.}%
+ \setcounter{eqfn}{\value{footnote}}%
+ \else%
+ \footnotemark[\value{eqfn}]%
+ \fi%
+ }%
+ \hsize\textwidth%
+ \linewidth\hsize%
+ \vskip 0.625in minus 0.125in%
+ \centering%
+ {\LARGE\bf \@title \par}%
+ \vskip 0.1in plus 0.5fil minus 0.05in%
+ {\Large{\textbf{\theauthors\ifhmode\\\fi}}}%
+ \vskip .2em plus 0.25fil%
+ {\normalsize \affiliations_\ifhmode\\\fi}%
+ \vskip .5em plus 2fil%
+ }%
+}%
+%
+\renewenvironment{abstract}{%
+ \centerline{\bf Abstract}%
+ \vspace{0.5ex}%
+ \setlength{\leftmargini}{10pt}%
+ \begin{quote}%
+ \small%
+}{%
+ \par%
+ \end{quote}%
+ \vskip 1ex%
+}%
+% jsp added:
+\def\pubnote#1{
+ \thispagestyle{myheadings}%
+ \pagestyle{myheadings}%
+ \markboth{#1}{#1}%
+ \setlength\headheight{10pt}%
+ \setlength\headsep{10pt}%
+}%
+%
+% SECTIONS with less space
+\def\section{\@startsection {section}{1}{\z@}{-2.0ex plus
+-0.5ex minus -.2ex}{3pt plus 2pt minus 1pt}{\Large\bf\centering}}
+\def\subsection{\@startsection{subsection}{2}{\z@}{-2.0ex plus
+-0.5ex minus -.2ex}{3pt plus 2pt minus 1pt}{\large\bf\raggedright}}
+\def\subsubsection{\@startsection{subparagraph}{3}{\z@}{-6pt plus
+%%% DIEGO changed: 29/11/2009
+%% 2pt minus 1pt}{-1em}{\normalsize\bf}}
+-2pt minus -1pt}{-1em}{\normalsize\bf}}
+%%% END changed
+\renewcommand\paragraph{\@startsection{paragraph}{4}{\z@}{-6pt plus -2pt minus -1pt}{-1em}{\normalsize\bf}}%
+\setcounter{secnumdepth}{0}
+% add period to section (but not subsection) numbers, reduce space after
+%\renewcommand{\thesection}
+% {\arabic{section}.\hskip-0.6em}
+%\renewcommand{\thesubsection}
+% {\arabic{section}.\arabic{subsection}\hskip-0.6em}
+% FOOTNOTES
+\footnotesep 6.65pt %
+\skip\footins 9pt plus 4pt minus 2pt
+\def\footnoterule{\kern-3pt \hrule width 5pc \kern 2.6pt }
+\setcounter{footnote}{0}
+% LISTS AND PARAGRAPHS
+\parindent 10pt
+\topsep 4pt plus 1pt minus 2pt
+\partopsep 1pt plus 0.5pt minus 0.5pt
+\itemsep 0.5pt plus 1pt minus 0.5pt
+\parsep 2pt plus 1pt minus 0.5pt
+\leftmargin 10pt \leftmargini 13pt \leftmarginii 10pt \leftmarginiii 5pt \leftmarginiv 5pt \leftmarginv 5pt \leftmarginvi 5pt
+\labelwidth\leftmargini\advance\labelwidth-\labelsep \labelsep 5pt
+\def\@listi{\leftmargin\leftmargini}
+\def\@listii{\leftmargin\leftmarginii
+\labelwidth\leftmarginii\advance\labelwidth-\labelsep
+\topsep 2pt plus 1pt minus 0.5pt
+\parsep 1pt plus 0.5pt minus 0.5pt
+\itemsep \parsep}
+\def\@listiii{\leftmargin\leftmarginiii
+\labelwidth\leftmarginiii\advance\labelwidth-\labelsep
+\topsep 1pt plus 0.5pt minus 0.5pt
+\parsep \z@
+\partopsep 0.5pt plus 0pt minus 0.5pt
+\itemsep \topsep}
+\def\@listiv{\leftmargin\leftmarginiv
+\labelwidth\leftmarginiv\advance\labelwidth-\labelsep}
+\def\@listv{\leftmargin\leftmarginv
+\labelwidth\leftmarginv\advance\labelwidth-\labelsep}
+\def\@listvi{\leftmargin\leftmarginvi
+\labelwidth\leftmarginvi\advance\labelwidth-\labelsep}
+\abovedisplayskip 7pt plus2pt minus5pt%
+\belowdisplayskip \abovedisplayskip
+\abovedisplayshortskip 0pt plus3pt%
+\belowdisplayshortskip 4pt plus3pt minus3pt%
+% Less leading in most fonts (due to the narrow columns)
+% The choices were between 1-pt and 1.5-pt leading
+\def\normalsize{\@setfontsize\normalsize\@xpt{11}} % 10 point on 11
+\def\small{\@setfontsize\small\@ixpt{10}} % 9 point on 10
+\def\footnotesize{\@setfontsize\footnotesize\@ixpt{10}} % 9 point on 10
+\def\scriptsize{\@setfontsize\scriptsize\@viipt{10}} % 7 point on 8
+\def\tiny{\@setfontsize\tiny\@vipt{7}} % 6 point on 7
+\def\large{\@setfontsize\large\@xipt{12}} % 11 point on 12
+\def\Large{\@setfontsize\Large\@xiipt{14}} % 12 point on 14
+\def\LARGE{\@setfontsize\LARGE\@xivpt{16}} % 14 point on 16
+\def\huge{\@setfontsize\huge\@xviipt{20}} % 17 point on 20
+\def\Huge{\@setfontsize\Huge\@xxpt{23}} % 20 point on 23
+
+\AtBeginDocument{%
+ \@ifpackageloaded{natbib}%
+ {%
+ % When natbib is in use, set the proper style and fix a few things
+ \let\cite\citep
+ \let\shortcite\citeyearpar
+ \setcitestyle{aysep={}}
+ \setlength\bibhang{0pt}
+ \bibliographystyle{aaai24}
+ }{}%
+ \@ifpackageloaded{hyperref}%
+ {%
+ \PackageError{aaai}{You must not use hyperref in AAAI papers.}{You (or one of the packages you imported) are importing the hyperref package, which is forbidden in AAAI papers. You must remove it from the paper to proceed.}
+ }{}%
+ \@ifpackageloaded{bbm}%
+ {%
+ \PackageError{aaai}{You must not use bbm package in AAAI papers because it introduces Type 3 fonts which are forbidden.}{See https://tex.stackexchange.com/questions/479160/a-replacement-to-mathbbm1-with-type-1-fonts for possible alternatives.}
+ }{}%
+ \@ifpackageloaded{authblk}%
+ {%
+ \PackageError{aaai}{Package authblk is forbbidden.}{Package authblk is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{balance}%
+ {%
+ \PackageError{aaai}{Package balance is forbbidden.}{Package balance is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{CJK}%
+ {%
+ \PackageError{aaai}{Package CJK is forbbidden.}{Package CJK is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{flushend}%
+ {%
+ \PackageError{aaai}{Package flushend is forbbidden.}{Package flushend is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{fontenc}%
+ {%
+ \PackageError{aaai}{Package fontenc is forbbidden.}{Package fontenc is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{fullpage}%
+ {%
+ \PackageError{aaai}{Package fullpage is forbbidden.}{Package fullpage is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{geometry}%
+ {%
+ \PackageError{aaai}{Package geometry is forbbidden.}{Package geometry is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{grffile}%
+ {%
+ \PackageError{aaai}{Package grffile is forbbidden.}{Package grffile is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{navigator}%
+ {%
+ \PackageError{aaai}{Package navigator is forbbidden.}{Package navigator is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{savetrees}%
+ {%
+ \PackageError{aaai}{Package savetrees is forbbidden.}{Package savetrees is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{setspace}%
+ {%
+ \PackageError{aaai}{Package setspace is forbbidden.}{Package setspace is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{stfloats}%
+ {%
+ \PackageError{aaai}{Package stfloats is forbbidden.}{Package stfloats is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{tabu}%
+ {%
+ \PackageError{aaai}{Package tabu is forbbidden.}{Package tabu is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{titlesec}%
+ {%
+ \PackageError{aaai}{Package titlesec is forbbidden.}{Package titlesec is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{tocbibind}%
+ {%
+ \PackageError{aaai}{Package tocbibind is forbbidden.}{Package tocbibind is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{ulem}%
+ {%
+ \PackageError{aaai}{Package ulem is forbbidden.}{Package ulem is forbbiden. You must find an alternative.}
+ }{}%
+ \@ifpackageloaded{wrapfig}%
+ {%
+ \PackageError{aaai}{Package wrapfig is forbbidden.}{Package wrapfig is forbbiden. You must find an alternative.}
+ }{}%
+}
+
+\let\endthebibliography=\endlist
diff --git a/AAAI/anonymous-submission-latex-2024.pdf b/AAAI/anonymous-submission-latex-2024.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..6f0d15331aeaab0a2ad1348f08c9768014ae6ddb
Binary files /dev/null and b/AAAI/anonymous-submission-latex-2024.pdf differ
diff --git a/AAAI/anonymous-submission-latex-2024.tex b/AAAI/anonymous-submission-latex-2024.tex
new file mode 100644
index 0000000000000000000000000000000000000000..923385c016b20329cc2dd7ed5b8867e44a557ae4
--- /dev/null
+++ b/AAAI/anonymous-submission-latex-2024.tex
@@ -0,0 +1,431 @@
+%File: anonymous-submission-latex-2024.tex
+\documentclass[letterpaper]{article} % DO NOT CHANGE THIS
+\usepackage[submission]{aaai24} % DO NOT CHANGE THIS
+\usepackage{times} % DO NOT CHANGE THIS
+\usepackage{helvet} % DO NOT CHANGE THIS
+\usepackage{courier} % DO NOT CHANGE THIS
+\usepackage[hyphens]{url} % DO NOT CHANGE THIS
+\usepackage{graphicx} % DO NOT CHANGE THIS
+\urlstyle{rm} % DO NOT CHANGE THIS
+\def\UrlFont{\rm} % DO NOT CHANGE THIS
+\usepackage{natbib} % DO NOT CHANGE THIS AND DO NOT ADD ANY OPTIONS TO IT
+\usepackage{caption} % DO NOT CHANGE THIS AND DO NOT ADD ANY OPTIONS TO IT
+\frenchspacing % DO NOT CHANGE THIS
+\setlength{\pdfpagewidth}{8.5in} % DO NOT CHANGE THIS
+\setlength{\pdfpageheight}{11in} % DO NOT CHANGE THIS
+%
+% These are recommended to typeset algorithms but not required. See the subsubsection on algorithms. Remove them if you don't have algorithms in your paper.
+\usepackage{algorithm}
+\usepackage{algorithmic}
+
+%
+% These are are recommended to typeset listings but not required. See the subsubsection on listing. Remove this block if you don't have listings in your paper.
+\usepackage{newfloat}
+\usepackage{listings}
+\DeclareCaptionStyle{ruled}{labelfont=normalfont,labelsep=colon,strut=off} % DO NOT CHANGE THIS
+\lstset{%
+ basicstyle={\footnotesize\ttfamily},% footnotesize acceptable for monospace
+ numbers=left,numberstyle=\footnotesize,xleftmargin=2em,% show line numbers, remove this entire line if you don't want the numbers.
+ aboveskip=0pt,belowskip=0pt,%
+ showstringspaces=false,tabsize=2,breaklines=true}
+\floatstyle{ruled}
+\newfloat{listing}{tb}{lst}{}
+\floatname{listing}{Listing}
+%
+% Keep the \pdfinfo as shown here. There's no need
+% for you to add the /Title and /Author tags.
+\pdfinfo{
+/TemplateVersion (2024.1)
+}
+
+\usepackage[utf8]{inputenc} % allow utf-8 input
+\usepackage[T1]{fontenc} % use 8-bit T1 fonts
+\usepackage{hyperref} % hyperlinks
+\usepackage{url} % simple URL typesetting
+\usepackage{booktabs} % professional-quality tables
+\usepackage{amsfonts} % blackboard math symbols
+\usepackage{nicefrac} % compact symbols for 1/2, etc.
+\usepackage{microtype} % microtypography
+\usepackage{xcolor} % colors
+
+\usepackage{amsmath}
+\usepackage{amsthm}
+\usepackage{caption}
+\usepackage{graphicx}
+\usepackage{algpseudocode}
+\usepackage{import}
+\usepackage{booktabs}
+\usepackage{longtable}
+\usepackage{array}
+\usepackage{multirow}
+
+
+% Numbered Environments:
+\newtheorem{definition}{Definition}[section]
+\newtheorem{question}{Research Question}[section]
+
+% DISALLOWED PACKAGES
+% \usepackage{authblk} -- This package is specifically forbidden
+% \usepackage{balance} -- This package is specifically forbidden
+% \usepackage{color (if used in text)
+% \usepackage{CJK} -- This package is specifically forbidden
+% \usepackage{float} -- This package is specifically forbidden
+% \usepackage{flushend} -- This package is specifically forbidden
+% \usepackage{fontenc} -- This package is specifically forbidden
+% \usepackage{fullpage} -- This package is specifically forbidden
+% \usepackage{geometry} -- This package is specifically forbidden
+% \usepackage{grffile} -- This package is specifically forbidden
+% \usepackage{hyperref} -- This package is specifically forbidden
+% \usepackage{navigator} -- This package is specifically forbidden
+% (or any other package that embeds links such as navigator or hyperref)
+% \indentfirst} -- This package is specifically forbidden
+% \layout} -- This package is specifically forbidden
+% \multicol} -- This package is specifically forbidden
+% \nameref} -- This package is specifically forbidden
+% \usepackage{savetrees} -- This package is specifically forbidden
+% \usepackage{setspace} -- This package is specifically forbidden
+% \usepackage{stfloats} -- This package is specifically forbidden
+% \usepackage{tabu} -- This package is specifically forbidden
+% \usepackage{titlesec} -- This package is specifically forbidden
+% \usepackage{tocbibind} -- This package is specifically forbidden
+% \usepackage{ulem} -- This package is specifically forbidden
+% \usepackage{wrapfig} -- This package is specifically forbidden
+% DISALLOWED COMMANDS
+% \nocopyright -- Your paper will not be published if you use this command
+% \addtolength -- This command may not be used
+% \balance -- This command may not be used
+% \baselinestretch -- Your paper will not be published if you use this command
+% \clearpage -- No page breaks of any kind may be used for the final version of your paper
+% \columnsep -- This command may not be used
+% \newpage -- No page breaks of any kind may be used for the final version of your paper
+% \pagebreak -- No page breaks of any kind may be used for the final version of your paperr
+% \pagestyle -- This command may not be used
+% \tiny -- This is not an acceptable font size.
+% \vspace{- -- No negative value may be used in proximity of a caption, figure, table, section, subsection, subsubsection, or reference
+% \vskip{- -- No negative value may be used to alter spacing above or below a caption, figure, table, section, subsection, subsubsection, or reference
+
+\setcounter{secnumdepth}{0} %May be changed to 1 or 2 if section numbers are desired.
+
+% The file aaai24.sty is the style file for AAAI Press
+% proceedings, working notes, and technical reports.
+%
+
+% Title
+
+% Your title must be in mixed case, not sentence case.
+% That means all verbs (including short verbs like be, is, using,and go),
+% nouns, adverbs, adjectives should be capitalized, including both words in hyphenated terms, while
+% articles, conjunctions, and prepositions are lower case unless they
+% directly follow a colon or long dash
+\title{AAAI Press Anonymous Submission\\Instructions for Authors Using \LaTeX{}}
+\author{
+ %Authors
+ % All authors must be in the same font size and format.
+ Written by AAAI Press Staff\textsuperscript{\rm 1}\thanks{With help from the AAAI Publications Committee.}\\
+ AAAI Style Contributions by Pater Patel Schneider,
+ Sunil Issar,\\
+ J. Scott Penberthy,
+ George Ferguson,
+ Hans Guesgen,
+ Francisco Cruz\equalcontrib,
+ Marc Pujol-Gonzalez\equalcontrib
+}
+\affiliations{
+ %Afiliations
+ \textsuperscript{\rm 1}Association for the Advancement of Artificial Intelligence\\
+ % If you have multiple authors and multiple affiliations
+ % use superscripts in text and roman font to identify them.
+ % For example,
+
+ % Sunil Issar\textsuperscript{\rm 2},
+ % J. Scott Penberthy\textsuperscript{\rm 3},
+ % George Ferguson\textsuperscript{\rm 4},
+ % Hans Guesgen\textsuperscript{\rm 5}
+ % Note that the comma should be placed after the superscript
+
+ 1900 Embarcadero Road, Suite 101\\
+ Palo Alto, California 94303-3310 USA\\
+ % email address must be in roman text type, not monospace or sans serif
+ proceedings-questions@aaai.org
+%
+% See more examples next
+}
+
+%Example, Single Author, ->> remove \iffalse,\fi and place them surrounding AAAI title to use it
+\iffalse
+\title{My Publication Title --- Single Author}
+\author {
+ Author Name
+}
+\affiliations{
+ Affiliation\\
+ Affiliation Line 2\\
+ name@example.com
+}
+\fi
+
+\iffalse
+%Example, Multiple Authors, ->> remove \iffalse,\fi and place them surrounding AAAI title to use it
+\title{My Publication Title --- Multiple Authors}
+\author {
+ % Authors
+ First Author Name\textsuperscript{\rm 1},
+ Second Author Name\textsuperscript{\rm 2},
+ Third Author Name\textsuperscript{\rm 1}
+}
+\affiliations {
+ % Affiliations
+ \textsuperscript{\rm 1}Affiliation 1\\
+ \textsuperscript{\rm 2}Affiliation 2\\
+ firstAuthor@affiliation1.com, secondAuthor@affilation2.com, thirdAuthor@affiliation1.com
+}
+\fi
+
+
+% REMOVE THIS: bibentry
+% This is only needed to show inline citations in the guidelines document. You should not need it and can safely delete it.
+\usepackage{bibentry}
+% END REMOVE bibentry
+
+\begin{document}
+
+\maketitle
+
+\begin{abstract}
+ Counterfactual Explanations offer an intuitive and straightforward way to explain black-box models and offer Algorithmic Recourse to individuals. To address the need for plausible explanations, existing work has primarily relied on surrogate models to learn how the input data is distributed. This effectively reallocates the task of learning realistic explanations for the data from the model itself to the surrogate. Consequently, the generated explanations may seem plausible to humans but need not necessarily describe the behaviour of the black-box model faithfully. We formalise this notion of faithfulness through the introduction of a tailored evaluation metric and propose a novel algorithmic framework for generating \textbf{E}nergy-\textbf{C}onstrained \textbf{C}onformal \textbf{Co}unterfactuals (ECCCos) that are only as plausible as the model permits. Through extensive empirical studies, we demonstrate that ECCCos reconcile the need for faithfulness and plausibility. In particular, we show that for models with gradient access, it is possible to achieve state-of-the-art performance without the need for surrogate models. To do so, our framework relies solely on properties defining the black-box model itself by leveraging recent advances in Energy-Based Modelling and Conformal Prediction. To our knowledge, this is the first venture in this direction for generating faithful Counterfactual Explanations. Thus, we anticipate that ECCCos can serve as a baseline for future research. We believe that our work opens avenues for researchers and practitioners seeking tools to better distinguish trustworthy from unreliable models.
+ \end{abstract}
+
+ \section{Introduction}\label{intro}
+
+ Counterfactual Explanations (CE) provide a powerful, flexible and intuitive way to not only explain black-box models but also help affected individuals through the means of Algorithmic Recourse. Instead of opening the Black Box, CE works under the premise of strategically perturbing model inputs to understand model behaviour~\citep{wachter2017counterfactual}. Intuitively speaking, we generate explanations in this context by asking what-if questions of the following nature: `Our credit risk model currently predicts that this individual is not credit-worthy. What if they reduced their monthly expenditures by 10\%?'
+
+ This is typically implemented by defining a target outcome $\mathbf{y}^+ \in \mathcal{Y}$ for some individual $\mathbf{x} \in \mathcal{X}=\mathbb{R}^D$ described by $D$ attributes, for which the model $M_{\theta}:\mathcal{X}\mapsto\mathcal{Y}$ initially predicts a different outcome: $M_{\theta}(\mathbf{x})\ne \mathbf{y}^+$. Counterfactuals are then searched by minimizing a loss function that compares the predicted model output to the target outcome: $\text{yloss}(M_{\theta}(\mathbf{x}),\mathbf{y}^+)$. Since CE work directly with the black-box model, valid counterfactuals always have full local fidelity by construction where fidelity is defined as the degree to which explanations approximate the predictions of a black-box model~\citep{mothilal2020explaining,molnar2020interpretable}.
+
+ In situations where full fidelity is a requirement, CE offer a more appropriate solution to Explainable Artificial Intelligence (XAI) than other popular approaches like LIME~\citep{ribeiro2016why} and SHAP~\citep{lundberg2017unified}, which involve local surrogate models. But even full fidelity is not a sufficient condition for ensuring that an explanation faithfully describes the behaviour of a model. That is because multiple very distinct explanations can all lead to the same model prediction, especially when dealing with heavily parameterized models like deep neural networks, which are typically underspecified by the data~\citep{wilson2020case}.
+
+ In the context of CE, the idea that no two explanations are the same arises almost naturally. A key focus in the literature has therefore been to identify those explanations and algorithmic recourses that are most appropriate based on a myriad of desiderata such as sparsity, actionability and plausibility. In this work, we draw closer attention to model faithfulness rather than fidelity as a desideratum for counterfactuals. Our key contributions are as follows:
+
+ \begin{itemize}
+ \item We show that fidelity is an insufficient evaluation metric for counterfactuals (Section~\ref{fidelity}) and propose a definition of faithfulness that gives rise to more suitable metrics (Section~\ref{faithfulness}).
+ \item We introduce a novel algorithmic approach for generating Energy-Constrained Conformal Counterfactuals (ECCCos) in Section~\ref{meth}.
+ \item We provide extensive empirical evidence demonstrating that ECCCos faithfully explain model behaviour and attain plausibility only when appropriate (Section~\ref{emp}).
+ \end{itemize}
+
+ To our knowledge, this is the first venture in this direction for generating faithful counterfactuals. Thus, we anticipate that ECCCos can serve as a baseline for future research. We believe that our work opens avenues for researchers and practitioners seeking tools to better distinguish trustworthy from unreliable models.
+
+ \section{Background}\label{background}
+
+ While CE can also be generated for arbitrary regression models~\citep{spooner2021counterfactual}, existing work has primarily focused on classification problems. Let $\mathcal{Y}=(0,1)^K$ denote the one-hot-encoded output domain with $K$ classes. Then most counterfactual generators rely on gradient descent to optimize different flavours of the following counterfactual search objective:
+
+ \begin{equation} \label{eq:general}
+ \begin{aligned}
+ \mathbf{Z}^\prime &= \arg \min_{\mathbf{Z}^\prime \in \mathcal{Z}^L} \left\{ {\text{yloss}(M_{\theta}(f(\mathbf{Z}^\prime)),\mathbf{y}^+)}+ \lambda {\text{cost}(f(\mathbf{Z}^\prime)) } \right\}
+ \end{aligned}
+ \end{equation}
+
+ Here $\text{yloss}(\cdot)$ denotes the primary loss function, $f(\cdot)$ is a function that maps from the counterfactual state space to the feature space and $\text{cost}(\cdot)$ is either a single penalty or a collection of penalties that are used to impose constraints through regularization. Equation~\ref{eq:general} restates the baseline approach to gradient-based counterfactual search proposed by~\citet{wachter2017counterfactual} in general form as introduced by~\citet{altmeyer2023endogenous}. To explicitly account for the multiplicity of explanations, $\mathbf{Z}^\prime=\{ \mathbf{z}_l\}_L$ denotes an $L$-dimensional array of counterfactual states.
+
+ The baseline approach, which we will simply refer to as \textit{Wachter}, searches a single counterfactual directly in the feature space and penalises its distance to the original factual. In this case, $f(\cdot)$ is simply the identity function and $\mathcal{Z}$ corresponds to the feature space itself. Many derivative works of~\citet{wachter2017counterfactual} have proposed new flavours of Equation~\ref{eq:general}, each of them designed to address specific \textit{desiderata} that counterfactuals ought to meet in order to properly serve both AI practitioners and individuals affected by algorithmic decision-making systems. The list of desiderata includes but is not limited to the following: sparsity, proximity~\citep{wachter2017counterfactual}, actionability~\citep{ustun2019actionable}, diversity~\citep{mothilal2020explaining}, plausibility~\citep{joshi2019realistic,poyiadzi2020face,schut2021generating}, robustness~\citep{upadhyay2021robust,pawelczyk2022probabilistically,altmeyer2023endogenous} and causality~\citep{karimi2021algorithmic}. Different counterfactual generators addressing these needs have been extensively surveyed and evaluated in various studies~\citep{verma2020counterfactual,karimi2020survey,pawelczyk2021carla,artelt2021evaluating,guidotti2022counterfactual}.
+
+ Perhaps unsurprisingly, the different desiderata are often positively correlated. For example, \citet{artelt2021evaluating} find that plausibility typically also leads to improved robustness. Similarly, plausibility has also been connected to causality in the sense that plausible counterfactuals respect causal relationships~\citep{mahajan2020preserving}. Consequently, the plausibility of counterfactuals has been among the primary concerns for researchers. Achieving plausibility is equivalent to ensuring that the generated counterfactuals comply with the true and unobserved data-generating process (DGP). We define plausibility formally in this work as follows:
+
+ \begin{definition}[Plausible Counterfactuals]
+ \label{def:plausible}
+ Let $\mathcal{X}|\mathbf{y}^+= p(\mathbf{x}|\mathbf{y}^+)$ denote the true conditional distribution of samples in the target class $\mathbf{y}^+$. Then for $\mathbf{x}^{\prime}$ to be considered a plausible counterfactual, we need: $\mathbf{x}^{\prime} \sim \mathcal{X}|\mathbf{y}^+$.
+ \end{definition}
+
+ To generate plausible counterfactuals, we need to be able to quantify the DGP: $\mathcal{X}|\mathbf{y}^+$. One straightforward way to do this is to use surrogate models for the task. \citet{joshi2019realistic}, for example, suggest that instead of searching counterfactuals in the feature space $\mathcal{X}$, we can instead traverse a latent embedding $\mathcal{Z}$ (Equation~\ref{eq:general}) that implicitly codifies the DGP. To learn the latent embedding, they propose using a generative model such as a Variational Autoencoder (VAE). Provided the surrogate model is well-specified, their proposed approach called \textit{REVISE} can yield plausible explanations. Others have proposed similar approaches: \citet{dombrowski2021diffeomorphic} traverse the base space of a normalizing flow to solve Equation~\ref{eq:general}; \citet{poyiadzi2020face} use density estimators ($\hat{p}: \mathcal{X} \mapsto [0,1]$) to constrain the counterfactuals to dense regions in the feature space; and, finally, \citet{karimi2021algorithmic} assume knowledge about the structural causal model that generates the data.
+
+ A competing approach towards plausibility that is also closely related to this work instead relies on the black-box model itself. \citet{schut2021generating} show that to meet the plausibility objective we need not explicitly model the input distribution. Pointing to the undesirable engineering overhead induced by surrogate models, they propose that we rely on the implicit minimisation of predictive uncertainty instead. Their proposed methodology, which we will refer to as \textit{Schut}, solves Equation~\ref{eq:general} by greedily applying Jacobian-Based Saliency Map Attacks (JSMA) in the feature space with cross-entropy loss and no penalty at all. The authors demonstrate theoretically and empirically that their approach yields counterfactuals for which the model $M_{\theta}$ predicts the target label $\mathbf{y}^+$ with high confidence. Provided the model is well-specified, these counterfactuals are plausible. This idea hinges on the assumption that the black-box model provides well-calibrated predictive uncertainty estimates.
+
+ \section{Why Fidelity is not Enough}\label{fidelity}
+
+ As discussed in the introduction, any valid counterfactual also has full fidelity by construction: solutions to Equation~\ref{eq:general} are considered valid as soon as the label predicted by the model matches the target class. So while fidelity always applies, counterfactuals that address the various desiderata introduced above can look vastly different from each other.
+
+ To demonstrate this with an example, we have trained a simple image classifier $M_{\theta}$ on the well-known \textit{MNIST} dataset~\citep{lecun1998mnist}: a Multi-Layer Perceptron (\textit{MLP}) with above 90 percent test accuracy. No measures have been taken to improve the model's adversarial robustness or its capacity for predictive uncertainty quantification. The far left panel of Figure ~\ref{fig:motiv} shows a random sample drawn from the dataset. The underlying classifier correctly predicts the label `nine' for this image. For the given factual image and model, we have used \textit{Wachter}, \textit{Schut} and \textit{REVISE} to generate one counterfactual each in the target class `seven'. The perturbed images are shown next to the factual image from left to right in Figure ~\ref{fig:motiv}. Captions on top of the individual images indicate the generator along with the predicted probability that the image belongs to the target class. In all three cases that probability is above 90 percent and yet the counterfactuals look very different from each other.
+
+ \begin{figure}
+ \centering
+ \includegraphics[width=0.8\textwidth]{../artifacts/results/images/mnist_motivation.png}
+ \caption{Counterfactuals for turning a 9 (nine) into a 7 (seven): original image (left); then from left to right the counterfactuals generated using \textit{Wachter}, \textit{Schut} and \textit{REVISE}.}\label{fig:motiv}
+ \end{figure}
+
+ Since \textit{Wachter} is only concerned with proximity, the generated counterfactual is almost indistinguishable from the factual. The approach by~\citet{schut2021generating} expects a well-calibrated model that can generate predictive uncertainty estimates. Since this is not the case, the generated counterfactual looks like an adversarial example. Finally, the counterfactual generated by \textit{REVISE} looks much more plausible than the other two. But is it also more faithful to the behaviour of our \textit{MNIST} classifier? That is much less clear because the surrogate used by \textit{REVISE} introduces friction: the generated explanations no longer depend exclusively on the black-box model itself.
+
+ So which of the counterfactuals most faithfully explains the behaviour of our image classifier? Fidelity cannot help us to make that judgement, because all of these counterfactuals have full fidelity. Thus, fidelity is an insufficient evaluation metric to assess the faithfulness of CE.
+
+ \section{A New Notion of Faithfulness}\label{faithfulness}
+
+ Considering the limitations of fidelity as demonstrated in the previous section, analogous to Definition~\ref{def:plausible}, we introduce a new notion of faithfulness in the context of CE:
+
+ \begin{definition}[Faithful Counterfactuals]
+ \label{def:faithful}
+ Let $\mathcal{X}_{\theta}|\mathbf{y}^+ = p_{\theta}(\mathbf{x}|\mathbf{y}^+)$ denote the conditional distribution of $\mathbf{x}$ in the target class $\mathbf{y}^+$, where $\theta$ denotes the parameters of model $M_{\theta}$. Then for $\mathbf{x}^{\prime}$ to be considered a faithful counterfactual, we need: $\mathbf{x}^{\prime} \sim \mathcal{X}_{\theta}|\mathbf{y}^+$.
+ \end{definition}
+
+ In doing this, we merge in and nuance the concept of plausibility (Definition~\ref{def:plausible}) where the notion of `consistent with the data' becomes `consistent with what the model has learned about the data'.
+
+ \subsection{Quantifying the Model's Generative Property}
+
+ To assess counterfactuals with respect to Definition~\ref{def:faithful}, we need a way to quantify the posterior conditional distribution $p_{\theta}(\mathbf{x}|\mathbf{y}^+)$. To this end, we draw on recent advances in Energy-Based Modelling (EBM), a subdomain of machine learning that is concerned with generative or hybrid modelling~\citep{grathwohl2020your,du2020implicit}. In particular, note that if we fix $\mathbf{y}$ to our target value $\mathbf{y}^+$, we can conditionally draw from $p_{\theta}(\mathbf{x}|\mathbf{y}^+)$ by randomly initializing $\mathbf{x}_0$ and then using Stochastic Gradient Langevin Dynamics (SGLD) as follows,
+
+ \begin{equation}\label{eq:sgld}
+ \begin{aligned}
+ \mathbf{x}_{j+1} &\leftarrow \mathbf{x}_j - \frac{\epsilon^2}{2} \mathcal{E}(\mathbf{x}_j|\mathbf{y}^+) + \epsilon \mathbf{r}_j, && j=1,...,J
+ \end{aligned}
+ \end{equation}
+
+ where $\mathbf{r}_j \sim \mathcal{N}(\mathbf{0},\mathbf{I})$ is the stochastic term and the step-size $\epsilon$ is typically polynomially decayed~\citep{welling2011bayesian}. The term $\mathcal{E}(\mathbf{x}_j|\mathbf{y}^+)$ denotes the model energy conditioned on the target class label $\mathbf{y}^+$ which we specify as the negative logit corresponding to the target class label $\mathbf{y}^*$. To allow for faster sampling, we follow the common practice of choosing the step-size $\epsilon$ and the standard deviation of $\mathbf{r}_j$ separately. While $\mathbf{x}_J$ is only guaranteed to distribute as $p_{\theta}(\mathbf{x}|\mathbf{y}^*)$ if $\epsilon \rightarrow 0$ and $J \rightarrow \infty$, the bias introduced for a small finite $\epsilon$ is negligible in practice \citep{murphy2023probabilistic,grathwohl2020your}. Appendix~\ref{app:jem} provides additional implementation details for any tasks related to energy-based modelling.
+
+ Generating multiple samples using SGLD thus yields an empirical distribution $\hat{\mathbf{X}}_{\theta,\mathbf{y}^+}$ that approximates what the model has learned about the input data. While in the context of EBM, this is usually done during training, we propose to repurpose this approach during inference in order to evaluate and generate faithful model explanations.
+
+ \subsection{Evaluating Plausibility and Faithfulness}
+
+ The parallels between our definitions of plausibility and faithfulness imply that we can also use similar evaluation metrics in both cases. Since existing work has focused heavily on plausibility, it offers a useful starting point. In particular,~\citet{guidotti2022counterfactual} have proposed an implausibility metric that measures the distance of the counterfactual from its nearest neighbour in the target class. As this distance is reduced, counterfactuals get more plausible under the assumption that the nearest neighbour itself is plausible in the sense of Definition~\ref{def:plausible}. In this work, we use the following adapted implausibility metric,
+
+ \begin{equation}\label{eq:impl}
+ \begin{aligned}
+ \text{impl}(\mathbf{x}^{\prime},\mathbf{X}_{\mathbf{y}^+}) = \frac{1}{\lvert\mathbf{X}_{\mathbf{y}^+}\rvert} \sum_{\mathbf{x} \in \mathbf{X}_{\mathbf{y}^+}} \text{dist}(\mathbf{x}^{\prime},\mathbf{x})
+ \end{aligned}
+ \end{equation}
+
+ where $\mathbf{x}^{\prime}$ denotes the counterfactual and $\mathbf{X}_{\mathbf{y}^+}$ is a subsample of the training data in the target class $\mathbf{y}^+$. By averaging over multiple samples in this manner, we avoid the risk that the nearest neighbour of $\mathbf{x}^{\prime}$ itself is not plausible according to Definition~\ref{def:plausible} (e.g an outlier).
+
+ Equation~\ref{eq:impl} gives rise to a similar evaluation metric for unfaithfulness. We merely swap out the subsample of individuals in the target class for a subset $\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}$ of the generated conditional samples:
+
+ \begin{equation}\label{eq:faith}
+ \begin{aligned}
+ \text{unfaith}(\mathbf{x}^{\prime},\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}) = \frac{1}{\lvert \hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}\rvert} \sum_{\mathbf{x} \in \hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}} \text{dist}(\mathbf{x}^{\prime},\mathbf{x})
+ \end{aligned}
+ \end{equation}
+
+ Specifically, we form this subset based on the $n_E$ generated samples with the lowest energy.
+
+ \section{Energy-Constrained Conformal Counterfactuals}\label{meth}
+
+ In this section, we describe \textit{ECCCo}, our proposed framework for generating Energy-Constrained Conformal Counterfactuals (ECCCos). It is based on the premise that counterfactuals should first and foremost be faithful. Plausibility, as a secondary concern, is then still attainable, but only to the degree that the black-box model itself has learned plausible explanations for the underlying data.
+
+ We begin by stating our proposed objective function, which involves tailored loss and penalty functions that we will explain in the following. In particular, we extend Equation~\ref{eq:general} as follows:
+
+ \begin{equation} \label{eq:eccco}
+ \begin{aligned}
+ \mathbf{Z}^\prime= \arg \min_{\mathbf{Z}^\prime \in \mathcal{Z}^M} &\{ {\text{yloss}(M_{\theta}(f(\mathbf{Z}^\prime)),\mathbf{y}^+)}+ \lambda_{1} {\text{dist}(f(\mathbf{Z}^\prime),\mathbf{x}) } \\
+ &+ \lambda_2 \text{unfaith}(f(\mathbf{Z}^\prime),\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}) + \lambda_3 \Omega(C_{\theta}(f(\mathbf{Z}^\prime);\alpha)) \}
+ \end{aligned}
+ \end{equation}
+
+ The first penalty term involving $\lambda_1$ induces proximity like in~\citet{wachter2017counterfactual}. Our default choice for $\text{dist}(\cdot)$ is the L1 Norm due to its sparsity-inducing properties. The second penalty term involving $\lambda_2$ induces faithfulness by constraining the energy of the generated counterfactual where $\text{unfaith}(\cdot)$ corresponds to the metric defined in Equation~\ref{eq:faith}. The third and final penalty term involving $\lambda_3$ introduces a new concept: it ensures that the generated counterfactual is associated with low predictive uncertainty. As mentioned above,~\citet{schut2021generating} have shown that plausible counterfactuals can be generated implicitly through predictive uncertainty minimization. Unfortunately, this relies on the assumption that the model itself can provide predictive uncertainty estimates, which may be too restrictive in practice.
+
+ To relax this assumption, we leverage recent advances in Conformal Prediction (CP), an approach to predictive uncertainty quantification that has recently gained popularity~\citep{angelopoulos2021gentle,manokhin2022awesome}. Crucially for our intended application, CP is model-agnostic and can be applied during inference without placing any restrictions on model training. Intuitively, CP works under the premise of turning heuristic notions of uncertainty into rigorous uncertainty estimates by repeatedly sifting through the training data or a dedicated calibration dataset. Conformal classifiers produce prediction sets for individual inputs that include all output labels that can be reasonably attributed to the input. These sets tend to be larger for inputs that do not conform with the training data and are characterized by high predictive uncertainty.
+
+ In order to generate counterfactuals that are associated with low predictive uncertainty, we use a smooth set size penalty introduced by~\citet{stutz2022learning} in the context of conformal training:
+
+ \begin{equation}\label{eq:setsize}
+ \begin{aligned}
+ \Omega(C_{\theta}(\mathbf{x};\alpha))&=\max \left(0, \sum_{\mathbf{y}\in\mathcal{Y}}C_{\theta,\mathbf{y}}(\mathbf{x}_i;\alpha) - \kappa \right)
+ \end{aligned}
+ \end{equation}
+
+ Here, $\kappa \in \{0,1\}$ is a hyper-parameter and $C_{\theta,\mathbf{y}}(\mathbf{x}_i;\alpha)$ can be interpreted as the probability of label $\mathbf{y}$ being included in the prediction set. In order to compute this penalty for any black-box model we merely need to perform a single calibration pass through a holdout set $\mathcal{D}_{\text{cal}}$. Arguably, data is typically abundant and in most applications, practitioners tend to hold out a test data set anyway. Consequently, CP removes the restriction on the family of predictive models, at the small cost of reserving a subset of the available data for calibration. This particular case of conformal prediction is referred to as Split Conformal Prediction (SCP) as it involves splitting the training data into a proper training dataset and a calibration dataset. In addition to the smooth set size penalty, we have also experimented with the use of a tailored function for $\text{yloss}(\cdot)$ that enforces that only the target label $\mathbf{y}^+$ is included in the prediction set ~\citet{stutz2022learning}. Further details are provided in Appendix~\ref{app:cp}.
+
+ \begin{figure}
+ \centering
+ \includegraphics[width=1.0\textwidth]{../artifacts/results/images/poc_gradient_fields.png}
+ \caption{Gradient fields and counterfactual paths for different generators. The objective is to generate a counterfual in the `blue' class for a sample from the `orange' class. Bright yellow stars indicate conditional samples generated through SGLD. The underlying classifier is a Joint Energy Model.}\label{fig:poc}
+ \end{figure}
+
+ \renewcommand{\algorithmicrequire}{\textbf{Input:}}
+ \renewcommand{\algorithmicensure}{\textbf{Output:}}
+
+ \begin{algorithm}
+ \caption{The \textit{ECCCo} generator}\label{alg:eccco}
+ \begin{algorithmic}[1]
+ \Require $\mathbf{x}, \mathbf{y}^+, M_{\theta}, f, \Lambda=[\lambda_1,\lambda_2,\lambda_3], \alpha, \mathcal{D}, T, \eta, n_{\mathcal{B}}, n_E$ where $M_{\theta}(\mathbf{x})\neq\mathbf{y}^+$
+ \Ensure $\mathbf{x}^\prime$
+ \State Initialize $\mathbf{z}^\prime \gets f^{-1}(\mathbf{x})$ \Comment{Map to counterfactual state space.}
+ \State Generate $\left\{\hat{\mathbf{x}}_{\theta,\mathbf{y}^+}\right\}_{n_{\mathcal{B}}} \gets p_{\theta}(\mathbf{x}_{\mathbf{y}^+})$ \Comment{Generate $n_{\mathcal{B}}$ samples using SGLD (Equation~\ref{eq:sgld}).}
+ \State Store $\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+} \gets \left\{\hat{\mathbf{x}}_{\theta,\mathbf{y}^+}\right\}_{n_{\mathcal{B}}}$ \Comment{Choose $n_E$ lowest-energy samples.}
+ \State Run \textit{SCP} for $M_{\theta}$ using $\mathcal{D}$ \Comment{Calibrate model through Split Conformal Prediction.}
+ \State Initialize $t \gets 0$
+ \While{\textit{not converged} or $t < T$} \Comment{For convergence conditions see Appendix~\ref{app:eccco}.}
+ \State $\mathbf{z}^\prime \gets \mathbf{z}^\prime - \eta \nabla_{\mathbf{z}^\prime} \mathcal{L}(\mathbf{z}^\prime,\mathbf{y}^+,\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}; \Lambda, \alpha)$ \Comment{Take gradient step of size $\eta$.}
+ \State $t \gets t+1$
+ \EndWhile
+ \State $\mathbf{x}^\prime \gets f(\mathbf{z}^\prime)$ \Comment{Map back to feature space.}
+ \end{algorithmic}
+ \end{algorithm}
+
+ To provide some further intuition about our objective defined in Equation~\ref{eq:eccco}, Figure~\ref{fig:poc} illustrates how the different components affect the counterfactual search for a synthetic dataset. The underlying classifier is a Joint Energy Model (\textit{JEM}) that was trained to predict the output class (`blue' or `orange') and generate class-conditional samples~\citep{grathwohl2020your}. We have used four different generator flavours to produce a counterfactual in the `blue' class for a sample from the `orange' class: \textit{Wachter}, which only uses the first penalty ($\lambda_2=\lambda_3=0$); \textit{ECCCo (no EBM)}, which does not constrain energy ($\lambda_2=0$); \textit{ECCCo (no CP)}, which involves no set size penalty ($\lambda_3=0$); and, finally, \textit{ECCCo}, which involves all penalties defined in Equation~\ref{eq:eccco}. Arrows indicate (negative) gradients with respect to the objective function at different points in the feature space.
+
+ While \textit{Wachter} generates a valid counterfactual, it ends up close to the original starting point consistent with its objective. \textit{ECCCo (no EBM)} pushes the counterfactual further into the target domain to minimize predictive uncertainty, but the outcome is still not plausible. The counterfactual produced by \textit{ECCCo (no CP)} is attracted by the generated samples shown in bright yellow. Since the \textit{JEM} has learned the conditional input distribution reasonably well in this case, the counterfactuals are both faithful and plausible. Finally, the outcome for \textit{ECCCo} looks similar, but the additional smooth set size penalty leads to somewhat faster convergence.
+
+ Algorithm~\ref{alg:eccco} describes how exactly \textit{ECCCo} works. For the sake of simplicity and without loss of generality, we limit our attention to generating a single counterfactual $\mathbf{x}^\prime=f(\mathbf{z}^\prime)$. The counterfactual state $\mathbf{z}^\prime$ is initialized by passing the factual $\mathbf{x}$ through a simple feature transformer $f^{-1}$. Next, we generate $n_{\mathcal{B}}$ conditional samples $\hat{\mathbf{x}}_{\theta,\mathbf{y}^+}$ using SGLD (Equation~\ref{eq:sgld}) and store the $n_E$ instances with the lowest energy. We then calibrate the model $M_{\theta}$ through Split Conformal Prediction. Finally, we search counterfactuals through gradient descent where $\mathcal{L}(\mathbf{z}^\prime,\mathbf{y}^+,\hat{\mathbf{X}}^{n_E}_{\theta,\mathbf{y}^+}; \Lambda, \alpha)$ denotes our loss function defined in Equation~\ref{eq:eccco}. The search terminates once the convergence criterium is met or the maximum number of iterations $T$ has been exhausted. Note that the choice of convergence criterium has important implications on the final counterfactual which we explain in Appendix~\ref{app:eccco}.
+
+ \section{Empirical Analysis}\label{emp}
+
+ Our goal in this section is to shed light on the following research questions:
+
+ \begin{question}[Faithfulness]\label{rq:faithfulness}
+ Are ECCCos more faithful than counterfactuals produced by our benchmark generators?
+ \end{question}
+
+ \begin{question}[Balancing Objectives]\label{rq:plausibility}
+ Compared to our benchmark generators, how do ECCCos balance the two key objectives of faithfulness and plausibility?
+ \end{question}
+
+ The second question is motivated by the intuition that faithfulness and plausibility should coincide for models that have learned plausible explanations of the data. Next, we first briefly describe our experimental setup before presenting our main results.
+
+ \subsection{Experimental Setup}
+
+ To assess and benchmark the performance of our proposed generator against the state of the art, we generate multiple counterfactuals for different models and datasets. In particular, we compare \textit{ECCCo} and its variants to the following counterfactual generators that were introduced above: firstly; \textit{Schut}, which works under the premise of minimizing predictive uncertainty; secondly, \textit{REVISE}, which is state-of-the-art with respect to plausibility; and, finally, \textit{Wachter}, which serves as our baseline.
+
+ We use both synthetic and real-world datasets from different domains, all of which are publicly available and commonly used to train and benchmark classification algorithms. We synthetically generate a dataset containing two \textit{Linearly Separable} Gaussian clusters ($n=1000$), as well as the well-known \textit{Circles} ($n=1000$) and \textit{Moons} ($n=2500$) data. Since these data are generated by distributions of varying degrees of complexity, they allow us to assess how the generators and our proposed evaluation metrics handle this.
+
+ As for real-world data, we follow~\citet{schut2021generating} and use the \textit{MNIST}~\citep{lecun1998mnist} dataset containing images of handwritten digits such as the example shown above in Figure~\ref{fig:motiv}. From the social sciences domain, we include Give Me Some Credit (\textit{GMSC})~\citep{kaggle2011give}: a tabular dataset that has been studied extensively in the literature on Algorithmic Recourse~\citep{pawelczyk2021carla}. It consists of 11 numeric features that can be used to predict the binary outcome variable indicating whether retail borrowers experience financial distress.
+
+ For the predictive modelling tasks, we use simple neural networks (\textit{MLP}) and Joint Energy Models (\textit{JEM}). For the more complex real-world datasets we also use ensembling in each case. Both joint-energy modelling and ensembling have been associated with improved generative properties and adversarial robustness~\citep{grathwohl2020your,lakshminarayanan2016simple}, so we expect this to be positively correlated with the plausibility of ECCCos. To account for stochasticity, we generate multiple counterfactuals for each target class, generator, model and dataset. Specifically, we randomly sample $n^{-}$ times from the subset of individuals for which the given model predicts the non-target class $\mathbf{y}^{-}$ given the current target. We set $n^{-}=25$ for all of our synthetic datasets, $n^{-}=10$ for \textit{GMSC} and $n^{-}=5$ for \textit{MNIST}. Full details concerning our parameter choices, training procedures and model performance can be found in Appendix~\ref{app:setup}.
+
+ \subsection{Results for Synthetic Data}
+
+ Table~\ref{tab:results-synthetic} shows the key results for the synthetic datasets separated by model (first column) and generator (second column). The numerical columns show sample averages and standard deviations of our key evaluation metrics computed across all counterfactuals. We have highlighted the best outcome for each model and metric in bold. To provide some sense of effect sizes, we have added asterisks to indicate that a given value is at least one ($*$) or two ($**$) standard deviations lower than the baseline (\textit{Wachter}).
+
+ Starting with the high-level results for our \textit{Linearly Separable} data, we find that \textit{ECCCo} produces the most faithful counterfactuals for both black-box models. This is consistent with our design since \textit{ECCCo} directly enforces faithfulness through regularization. Crucially though, \textit{ECCCo} also produces the most plausible counterfactuals for both models. This dataset is so simple that even the \textit{MLP} has learned plausible explanations of the input data. Zooming in on the granular details for the \textit{Linearly Separable} data, the results for \textit{ECCCo (no CP)} and \textit{ECCCo (no EBM)} indicate that the positive results are dominated by the effect of quantifying and leveraging the model's generative property (EBM). Conformal Prediction alone only leads to marginally improved faithfulness and plausibility.
+
+ The findings for the \textit{Moons} dataset are broadly in line with the findings so far: for the \textit{JEM}, \textit{ECCCo} yields substantially more faithful and plausible counterfactuals than all other generators. For the \textit{MLP}, faithfulness is maintained but counterfactuals are not plausible. This high-level pattern is broadly consistent with other more complex datasets and supportive of our narrative, so it is worth highlighting: ECCCos consistently achieve high faithfulness, which---subject to the quality of the model itself---coincides with high plausibility. By comparison, \textit{REVISE} yields the most plausible counterfactuals for the \textit{MLP}, but it does so at the cost of faithfulness. We also observe that the best results for \textit{ECCCo} are achieved when using both penalties. Once again though, the generative component (EBM) has a stronger impact on the positive results for the \textit{JEM}.
+
+ For the \textit{Circles} data, it appears that \textit{REVISE} performs well, but we note that it generates valid counterfactuals only half of the time (see Appendix~\ref{app:results} for a complete overview including additional common evaluation metrics). The underlying VAE with default parameters has not adequately learned the data-generating process. Of course, it is possible to improve generative performance through hyperparameter tuning but this example serves to illustrate that \textit{REVISE} depends on the quality of its surrogate. Independent of the outcome for \textit{REVISE}, however, the results do not seem to indicate that \textit{ECCCo} substantially improves faithfulness and plausibility for the \textit{Circles} data. We think this points to a limitation of our evaluation metrics rather than \textit{ECCCo} itself: computing average distances fails to account for the `wraparound' effect associated with circular data~\citep{gill2010circular}.
+
+ \import{contents/}{table-synthetic.tex}
+
+ \subsection{Results for Real-World Data}
+
+ The results for our real-world datasets are shown in Table~\ref{tab:results-real-world}. Once again the findings indicate that the plausibility of ECCCos is positively correlated with the capacity of the black-box model to distinguish plausible from implausible inputs. The case is very clear for \textit{MNIST}: ECCCos are consistently more faithful than the counterfactuals produced by our benchmark generators and their plausibility gradually improves through ensembling and joint-energy modelling. Interestingly, faithfulness also gradually improves for \textit{REVISE}. This indicates that as our models improve, their generative capacity approaches that of the surrogate VAE used by \textit{REVISE}. The VAE still outperforms our classifiers in this regard, as evident from the fact that \textit{ECCCo} never quite reaches the same level of plausibility as \textit{REVISE}. With reference to Appendix~\ref{app:results} we note that the results for \textit{Schut} need to be discounted as it rarely produces valid counterfactuals for \textit{MNIST}. Relatedly, we find that \textit{ECCCo} is the only generator that consistently achieves full validity. Finally, it is worth noting that \textit{ECCCo} produces counterfactual images with the lowest average predictive uncertainty for all models.
+
+ For the tabular credit dataset (\textit{GMSC}) it is inherently challenging to use deep neural networks in order to achieve good discriminative performance~\citep{borisov2021deep,grinsztajn2022why} and generative performance~\citep{liu2023goggle}, respectively. In order to achieve high plausibility, \textit{ECCCo} effectively requires classifiers to achieve good performance for both tasks. Since this is a challenging task even for Joint Energy Models, it is not surprising to find that even though \textit{ECCCo} once again achieves state-of-the-art faithfulness, it is outperformed by \textit{REVISE} and \textit{Schut} with respect to plausibility.
+
+ \subsection{Key Takeways}
+
+ To conclude this section, we summarize our findings with reference to the opening questions. The results clearly demonstrate that \textit{ECCCo} consistently achieves state-of-the-art faithfulness, as it was designed to do (Research Question~\ref{rq:faithfulness}). A related important finding is that \textit{ECCCo} yields highly plausible explanations provided that they faithfully describe model behaviour (Research Question~\ref{rq:plausibility}). \textit{ECCCo} achieves this result primarily by leveraging the model's generative property.
+
+ \import{contents/}{table-real-world.tex}
+
+ \section{Limitations}
+
+ Even though we have taken considerable measures to study our proposed methodology carefully, limitations can still be identified. In particular, we have found that the performance of \textit{ECCCo} is sensitive to hyperparameter choices. In order to achieve faithfulness, we generally had to penalise the distance from generated samples slightly more than the distance from factual values.
+
+ Conversely, we have not found that strongly penalising prediction set sizes had any discernable effect. Our results indicate that CP alone is often not sufficient to achieve faithfulness and plausibility, although we acknowledge that this needs to be investigated more thoroughly through future work.
+
+ While our approach is readily applicable to models with gradient access like deep neural networks, more work is needed to generalise it to other machine learning models such as decision trees. Relatedly, common challenges associated with Energy-Based Modelling including sensitivity to scale, training instabilities and sensitivity to hyperparameters also apply to \textit{ECCCo}.
+
+ \section{Conclusion}
+
+ This work leverages recent advances in Energy-Based Modelling and Conformal Prediction in the context of Explainable Artificial Intelligence. We have proposed a new way to generate counterfactuals that are maximally faithful to the black-box model they aim to explain. Our proposed generator, \textit{ECCCo}, produces plausible counterfactuals if and only if the black-box model itself has learned realistic explanations for the data, which we have demonstrated through rigorous empirical analysis. This should enable researchers and practitioners to use counterfactuals in order to discern trustworthy models from unreliable ones. While the scope of this work limits its generalizability, we believe that \textit{ECCCo} offers a solid baseline for future work on faithful Counterfactual Explanations.
+
+ \section{Acknowledgments}
+
+ Some of the members of TU Delft were partially funded by ICAI AI for Fintech Research, an ING — TU Delft
+ collaboration.
+
+ \bibliography{bib}
+
+\end{document}
diff --git a/AAAI/bib.bib b/AAAI/bib.bib
new file mode 100644
index 0000000000000000000000000000000000000000..daed3cba558bf28a0516e5454625b5b48ea016e8
--- /dev/null
+++ b/AAAI/bib.bib
@@ -0,0 +1,2837 @@
+@TechReport{kingma2017adam,
+ author = {Kingma, Diederik P. and Ba, Jimmy},
+ date = {2017-01},
+ institution = {arXiv},
+ title = {Adam: {A} {Method} for {Stochastic} {Optimization}},
+ doi = {10.48550/arXiv.1412.6980},
+ note = {arXiv:1412.6980 [cs] type: article},
+ url = {http://arxiv.org/abs/1412.6980},
+ urldate = {2023-05-17},
+ abstract = {We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. The method is straightforward to implement, is computationally efficient, has little memory requirements, is invariant to diagonal rescaling of the gradients, and is well suited for problems that are large in terms of data and/or parameters. The method is also appropriate for non-stationary objectives and problems with very noisy and/or sparse gradients. The hyper-parameters have intuitive interpretations and typically require little tuning. Some connections to related algorithms, on which Adam was inspired, are discussed. We also analyze the theoretical convergence properties of the algorithm and provide a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework. Empirical results demonstrate that Adam works well in practice and compares favorably to other stochastic optimization methods. Finally, we discuss AdaMax, a variant of Adam based on the infinity norm.},
+ annotation = {Comment: Published as a conference paper at the 3rd International Conference for Learning Representations, San Diego, 2015},
+ file = {arXiv Fulltext PDF:https\://arxiv.org/pdf/1412.6980.pdf:application/pdf},
+ keywords = {Computer Science - Machine Learning},
+ shorttitle = {Adam},
+}
+
+@TechReport{xiao2017fashion,
+ author = {Xiao, Han and Rasul, Kashif and Vollgraf, Roland},
+ date = {2017-09},
+ institution = {arXiv},
+ title = {Fashion-{MNIST}: a {Novel} {Image} {Dataset} for {Benchmarking} {Machine} {Learning} {Algorithms}},
+ doi = {10.48550/arXiv.1708.07747},
+ note = {arXiv:1708.07747 [cs, stat] type: article},
+ url = {http://arxiv.org/abs/1708.07747},
+ urldate = {2023-05-10},
+ abstract = {We present Fashion-MNIST, a new dataset comprising of 28x28 grayscale images of 70,000 fashion products from 10 categories, with 7,000 images per category. The training set has 60,000 images and the test set has 10,000 images. Fashion-MNIST is intended to serve as a direct drop-in replacement for the original MNIST dataset for benchmarking machine learning algorithms, as it shares the same image size, data format and the structure of training and testing splits. The dataset is freely available at https://github.com/zalandoresearch/fashion-mnist},
+ annotation = {Comment: Dataset is freely available at https://github.com/zalandoresearch/fashion-mnist Benchmark is available at http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/},
+ file = {:xiao2017fashion - Fashion MNIST_ a Novel Image Dataset for Benchmarking Machine Learning Algorithms.pdf:PDF},
+ keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Statistics - Machine Learning},
+ shorttitle = {Fashion-{MNIST}},
+}
+
+@Online{mw2023fidelity,
+ author = {Merriam-Webster},
+ title = {"Fidelity"},
+ url = {https://www.merriam-webster.com/dictionary/fidelity},
+ language = {en},
+ organization = {Merriam-Webster},
+ urldate = {2023-03-23},
+ abstract = {the quality or state of being faithful; accuracy in details : exactness; the degree to which an electronic device (such as a record player, radio, or television) accurately reproduces its effect (such as sound or picture)… See the full definition},
+}
+
+@InProceedings{altmeyer2023endogenous,
+ author = {Altmeyer, Patrick and Angela, Giovan and Buszydlik, Aleksander and Dobiczek, Karol and van Deursen, Arie and Liem, Cynthia},
+ booktitle = {First {IEEE} {Conference} on {Secure} and {Trustworthy} {Machine} {Learning}},
+ title = {Endogenous {Macrodynamics} in {Algorithmic} {Recourse}},
+ file = {:altmeyerendogenous - Endogenous Macrodynamics in Algorithmic Recourse.pdf:PDF},
+ year = {2023},
+}
+
+%% This BibTeX bibliography file was created using BibDesk.
+%% https://bibdesk.sourceforge.io/
+
+%% Created for Anonymous Author at 2022-12-13 12:58:22 +0100
+
+
+%% Saved with string encoding Unicode (UTF-8)
+
+
+
+@Article{abadie2002instrumental,
+ author = {Abadie, Alberto and Angrist, Joshua and Imbens, Guido},
+ title = {Instrumental Variables Estimates of the Effect of Subsidized Training on the Quantiles of Trainee Earnings},
+ number = {1},
+ pages = {91--117},
+ volume = {70},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica : journal of the Econometric Society},
+ shortjournal = {Econometrica},
+ year = {2002},
+}
+
+@Article{abadie2003economic,
+ author = {Abadie, Alberto and Gardeazabal, Javier},
+ title = {The Economic Costs of Conflict: {{A}} Case Study of the {{Basque Country}}},
+ number = {1},
+ pages = {113--132},
+ volume = {93},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {American economic review},
+ year = {2003},
+}
+
+@InProceedings{ackerman2021machine,
+ author = {Ackerman, Samuel and Dube, Parijat and Farchi, Eitan and Raz, Orna and Zalmanovici, Marcel},
+ booktitle = {2021 {{IEEE}}/{{ACM Third International Workshop}} on {{Deep Learning}} for {{Testing}} and {{Testing}} for {{Deep Learning}} ({{DeepTest}})},
+ title = {Machine {{Learning Model Drift Detection Via Weak Data Slices}}},
+ pages = {1--8},
+ publisher = {{IEEE}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Article{allen2017referencedependent,
+ author = {Allen, Eric J and Dechow, Patricia M and Pope, Devin G and Wu, George},
+ title = {Reference-Dependent Preferences: {{Evidence}} from Marathon Runners},
+ number = {6},
+ pages = {1657--1672},
+ volume = {63},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Management Science},
+ year = {2017},
+}
+
+@Article{altmeyer2018option,
+ author = {Altmeyer, Patrick and Grapendal, Jacob Daniel and Pravosud, Makar and Quintana, Gand Derry},
+ title = {Option Pricing in the {{Heston}} Stochastic Volatility Model: An Empirical Evaluation},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2018},
+}
+
+@Article{altmeyer2021deep,
+ author = {Altmeyer, Patrick and Agusti, Marc and Vidal-Quadras Costa, Ignacio},
+ title = {Deep {{Vector Autoregression}} for {{Macroeconomic Data}}},
+ url = {https://thevoice.bse.eu/wp-content/uploads/2021/07/ds21-project-agusti-et-al.pdf},
+ bdsk-url-1 = {https://thevoice.bse.eu/wp-content/uploads/2021/07/ds21-project-agusti-et-al.pdf},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Book{altmeyer2021deepvars,
+ author = {Altmeyer, Patrick},
+ title = {Deepvars: {{Deep Vector Autoregession}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Misc{altmeyer2022counterfactualexplanations,
+ author = {Altmeyer, Patrick},
+ title = {{{CounterfactualExplanations}}.Jl - a {{Julia}} Package for {{Counterfactual Explanations}} and {{Algorithmic Recourse}}},
+ url = {https://github.com/pat-alt/CounterfactualExplanations.jl},
+ bdsk-url-1 = {https://github.com/pat-alt/CounterfactualExplanations.jl},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2022},
+}
+
+@Software{altmeyerCounterfactualExplanationsJlJulia2022,
+ author = {Altmeyer, Patrick},
+ title = {{{CounterfactualExplanations}}.Jl - a {{Julia}} Package for {{Counterfactual Explanations}} and {{Algorithmic Recourse}}},
+ url = {https://github.com/pat-alt/CounterfactualExplanations.jl},
+ version = {0.1.2},
+ bdsk-url-1 = {https://github.com/pat-alt/CounterfactualExplanations.jl},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2022},
+}
+
+@Unpublished{angelopoulos2021gentle,
+ author = {Angelopoulos, Anastasios N. and Bates, Stephen},
+ title = {A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2107.07511},
+ eprinttype = {arxiv},
+ file = {:/Users/FA31DU/Zotero/storage/RKSUMYZG/Angelopoulos and Bates - 2021 - A gentle introduction to conformal prediction and .pdf:;:/Users/FA31DU/Zotero/storage/PRUEKRR3/2107.html:},
+ year = {2021},
+}
+
+@Misc{angelopoulos2022uncertainty,
+ author = {Angelopoulos, Anastasios and Bates, Stephen and Malik, Jitendra and Jordan, Michael I.},
+ title = {Uncertainty {{Sets}} for {{Image Classifiers}} Using {{Conformal Prediction}}},
+ eprint = {2009.14193},
+ eprinttype = {arxiv},
+ abstract = {Convolutional image classifiers can achieve high predictive accuracy, but quantifying their uncertainty remains an unresolved challenge, hindering their deployment in consequential settings. Existing uncertainty quantification techniques, such as Platt scaling, attempt to calibrate the network's probability estimates, but they do not have formal guarantees. We present an algorithm that modifies any classifier to output a predictive set containing the true label with a user-specified probability, such as 90\%. The algorithm is simple and fast like Platt scaling, but provides a formal finite-sample coverage guarantee for every model and dataset. Our method modifies an existing conformal prediction algorithm to give more stable predictive sets by regularizing the small scores of unlikely classes after Platt scaling. In experiments on both Imagenet and Imagenet-V2 with ResNet-152 and other classifiers, our scheme outperforms existing approaches, achieving coverage with sets that are often factors of 5 to 10 smaller than a stand-alone Platt scaling baseline.},
+ archiveprefix = {arXiv},
+ bdsk-url-1 = {http://arxiv.org/abs/2009.14193},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ file = {:/Users/FA31DU/Zotero/storage/5BYIRBR2/Angelopoulos et al. - 2022 - Uncertainty Sets for Image Classifiers using Confo.pdf:;:/Users/FA31DU/Zotero/storage/2QJAKFKV/2009.html:},
+ keywords = {Computer Science - Computer Vision and Pattern Recognition, Mathematics - Statistics Theory, Statistics - Machine Learning},
+ month = sep,
+ number = {arXiv:2009.14193},
+ primaryclass = {cs, math, stat},
+ publisher = {{arXiv}},
+ year = {2022},
+}
+
+@Article{angelucci2009indirect,
+ author = {Angelucci, Manuela and De Giorgi, Giacomo},
+ title = {Indirect Effects of an Aid Program: How Do Cash Transfers Affect Ineligibles' Consumption?},
+ number = {1},
+ pages = {486--508},
+ volume = {99},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {American economic review},
+ year = {2009},
+}
+
+@Article{angrist1990lifetime,
+ author = {Angrist, Joshua D},
+ title = {Lifetime Earnings and the {{Vietnam}} Era Draft Lottery: Evidence from Social Security Administrative Records},
+ pages = {313--336},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The American Economic Review},
+ year = {1990},
+}
+
+@Unpublished{antoran2020getting,
+ author = {Antor{\'a}n, Javier and Bhatt, Umang and Adel, Tameem and Weller, Adrian and Hern{\'a}ndez-Lobato, Jos{\'e} Miguel},
+ title = {Getting a Clue: {{A}} Method for Explaining Uncertainty Estimates},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2006.06848},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Article{arcones1992bootstrap,
+ author = {Arcones, Miguel A and Gine, Evarist},
+ title = {On the Bootstrap of {{U}} and {{V}} Statistics},
+ pages = {655--674},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Annals of Statistics},
+ year = {1992},
+}
+
+@Article{ariely2003coherent,
+ author = {Ariely, Dan and Loewenstein, George and Prelec, Drazen},
+ title = {``{{Coherent}} Arbitrariness'': {{Stable}} Demand Curves without Stable Preferences},
+ number = {1},
+ pages = {73--106},
+ volume = {118},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Quarterly journal of economics},
+ year = {2003},
+}
+
+@Article{ariely2006tom,
+ author = {Ariely, Dan and Loewenstein, George and Prelec, Drazen},
+ title = {Tom {{Sawyer}} and the Construction of Value},
+ number = {1},
+ pages = {1--10},
+ volume = {60},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Economic Behavior \& Organization},
+ year = {2006},
+}
+
+@Article{arrieta2020explainable,
+ author = {Arrieta, Alejandro Barredo and Diaz-Rodriguez, Natalia and Del Ser, Javier and Bennetot, Adrien and Tabik, Siham and Barbado, Alberto and Garcia, Salvador and Gil-Lopez, Sergio and Molina, Daniel and Benjamins, Richard and others},
+ title = {Explainable {{Artificial Intelligence}} ({{XAI}}): {{Concepts}}, Taxonomies, Opportunities and Challenges toward Responsible {{AI}}},
+ pages = {82--115},
+ volume = {58},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Information Fusion},
+ year = {2020},
+}
+
+@Article{auer2002finitetime,
+ author = {Auer, Peter and Cesa-Bianchi, Nicolo and Fischer, Paul},
+ title = {Finite-Time Analysis of the Multiarmed Bandit Problem},
+ number = {2},
+ pages = {235--256},
+ volume = {47},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Machine learning},
+ year = {2002},
+}
+
+@Article{barabasi2016network,
+ author = {Barab{\'a}si, Albert-L{\'a}szl{\'o}},
+ title = {Network {{Science}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Network Science},
+ year = {2016},
+}
+
+@Unpublished{bastounis2021mathematics,
+ author = {Bastounis, Alexander and Hansen, Anders C and Vla{\v c}i{\'c}, Verner},
+ title = {The Mathematics of Adversarial Attacks in {{AI}}--{{Why}} Deep Learning Is Unstable despite the Existence of Stable Neural Networks},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2109.06098},
+ eprinttype = {arxiv},
+ year = {2021},
+}
+
+@Article{bechara1997deciding,
+ author = {Bechara, Antoine and Damasio, Hanna and Tranel, Daniel and Damasio, Antonio R},
+ title = {Deciding Advantageously before Knowing the Advantageous Strategy},
+ number = {5304},
+ pages = {1293--1295},
+ volume = {275},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Science (New York, N.Y.)},
+ shortjournal = {Science},
+ year = {1997},
+}
+
+@Book{berlinet2011reproducing,
+ author = {Berlinet, Alain and Thomas-Agnan, Christine},
+ title = {Reproducing Kernel {{Hilbert}} Spaces in Probability and Statistics},
+ publisher = {{Springer Science \& Business Media}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2011},
+}
+
+@Misc{bernanke1990federal,
+ author = {Bernanke, Ben S},
+ title = {The Federal Funds Rate and the Channels of Monetary Transnission},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ publisher = {{National Bureau of Economic Research Cambridge, Mass., USA}},
+ year = {1990},
+}
+
+@Article{besbes2014stochastic,
+ author = {Besbes, Omar and Gur, Yonatan and Zeevi, Assaf},
+ title = {Stochastic Multi-Armed-Bandit Problem with Non-Stationary Rewards},
+ pages = {199--207},
+ volume = {27},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in neural information processing systems},
+ year = {2014},
+}
+
+@Article{bholat2020impact,
+ author = {Bholat, D and Gharbawi, M and Thew, O},
+ title = {The {{Impact}} of {{Covid}} on {{Machine Learning}} and {{Data Science}} in {{UK Banking}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Bank of England Quarterly Bulletin, Q4},
+ year = {2020},
+}
+
+@Book{bishop2006pattern,
+ author = {Bishop, Christopher M},
+ title = {Pattern Recognition and Machine Learning},
+ publisher = {{springer}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2006},
+}
+
+@Article{blaom2020mlj,
+ author = {Blaom, Anthony D. and Kiraly, Franz and Lienart, Thibaut and Simillides, Yiannis and Arenas, Diego and Vollmer, Sebastian J.},
+ title = {{{MLJ}}: {{A Julia}} Package for Composable Machine Learning},
+ doi = {10.21105/joss.02704},
+ issn = {2475-9066},
+ number = {55},
+ pages = {2704},
+ urldate = {2022-10-27},
+ volume = {5},
+ abstract = {Blaom et al., (2020). MLJ: A Julia package for composable machine learning. Journal of Open Source Software, 5(55), 2704, https://doi.org/10.21105/joss.02704},
+ bdsk-url-1 = {https://joss.theoj.org/papers/10.21105/joss.02704},
+ bdsk-url-2 = {https://doi.org/10.21105/joss.02704},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ file = {:/Users/FA31DU/Zotero/storage/7AY87FGP/Blaom et al. - 2020 - MLJ A Julia package for composable machine learni.pdf:;:/Users/FA31DU/Zotero/storage/D69YSMVF/joss.html:},
+ journal = {Journal of Open Source Software},
+ langid = {english},
+ month = nov,
+ shorttitle = {{{MLJ}}},
+ year = {2020},
+}
+
+@InProceedings{blundell2015weight,
+ author = {Blundell, Charles and Cornebise, Julien and Kavukcuoglu, Koray and Wierstra, Daan},
+ booktitle = {International Conference on Machine Learning},
+ title = {Weight Uncertainty in Neural Network},
+ pages = {1613--1622},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2015},
+}
+
+@Article{borch2022machine,
+ author = {Borch, Christian},
+ title = {Machine Learning, Knowledge Risk, and Principal-Agent Problems in Automated Trading},
+ pages = {101852},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Technology in Society},
+ year = {2022},
+}
+
+@Unpublished{borisov2021deep,
+ author = {Borisov, Vadim and Leemann, Tobias and Se{\ss}ler, Kathrin and Haug, Johannes and Pawelczyk, Martin and Kasneci, Gjergji},
+ title = {Deep Neural Networks and Tabular Data: {{A}} Survey},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2110.01889},
+ eprinttype = {arxiv},
+ year = {2021},
+}
+
+@Article{bramoulle2009identification,
+ author = {Bramoull{\'e}, Yann and Djebbari, Habiba and Fortin, Bernard},
+ title = {Identification of Peer Effects through Social Networks},
+ number = {1},
+ pages = {41--55},
+ volume = {150},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of econometrics},
+ year = {2009},
+}
+
+@Article{bramoulle2020peer,
+ author = {Bramoull{\'e}, Yann and Djebbari, Habiba and Fortin, Bernard},
+ title = {Peer Effects in Networks: {{A}} Survey},
+ pages = {603--629},
+ volume = {12},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Annual Review of Economics},
+ year = {2020},
+}
+
+@Unpublished{branco2015survey,
+ author = {Branco, Paula and Torgo, Luis and Ribeiro, Rita},
+ title = {A Survey of Predictive Modelling under Imbalanced Distributions},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1505.01658},
+ eprinttype = {arxiv},
+ year = {2015},
+}
+
+@Book{brock1991nonlinear,
+ author = {Brock, William Allen and Brock, William A and Hsieh, David Arthur and LeBaron, Blake Dean and Brock, William E},
+ title = {Nonlinear Dynamics, Chaos, and Instability: Statistical Theory and Economic Evidence},
+ publisher = {{MIT press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1991},
+}
+
+@InProceedings{buolamwini2018gender,
+ author = {Buolamwini, Joy and Gebru, Timnit},
+ booktitle = {Conference on Fairness, Accountability and Transparency},
+ title = {Gender Shades: {{Intersectional}} Accuracy Disparities in Commercial Gender Classification},
+ pages = {77--91},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2018},
+}
+
+@Unpublished{bussmann2020neural,
+ author = {Bussmann, Bart and Nys, Jannes and Latr{\'e}, Steven},
+ title = {Neural {{Additive Vector Autoregression Models}} for {{Causal Discovery}} in {{Time Series Data}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2010.09429},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Report{card1993minimum,
+ author = {Card, David and Krueger, Alan B},
+ title = {Minimum Wages and Employment: {{A}} Case Study of the Fast Food Industry in {{New Jersey}} and {{Pennsylvania}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ school = {{National Bureau of Economic Research}},
+ year = {1993},
+}
+
+@InProceedings{carlini2017evaluating,
+ author = {Carlini, Nicholas and Wagner, David},
+ booktitle = {2017 Ieee Symposium on Security and Privacy (Sp)},
+ title = {Towards Evaluating the Robustness of Neural Networks},
+ pages = {39--57},
+ publisher = {{IEEE}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2017},
+}
+
+@Article{carlisle2019racist,
+ author = {Carlisle, M.},
+ title = {Racist Data Destruction? - a {{Boston}} Housing Dataset Controversy},
+ url = {https://medium.com/@docintangible/racist-data-destruction-113e3eff54a8},
+ bdsk-url-1 = {https://medium.com/@docintangible/racist-data-destruction-113e3eff54a8},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2019},
+}
+
+@Article{carrell2009does,
+ author = {Carrell, Scott E and Fullerton, Richard L and West, James E},
+ title = {Does Your Cohort Matter? {{Measuring}} Peer Effects in College Achievement},
+ number = {3},
+ pages = {439--464},
+ volume = {27},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Labor Economics},
+ year = {2009},
+}
+
+@Article{carrell2013natural,
+ author = {Carrell, Scott E and Sacerdote, Bruce I and West, James E},
+ title = {From Natural Variation to Optimal Policy? {{The}} Importance of Endogenous Peer Group Formation},
+ number = {3},
+ pages = {855--882},
+ volume = {81},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica : journal of the Econometric Society},
+ shortjournal = {Econometrica},
+ year = {2013},
+}
+
+@Article{carrizosa2021generating,
+ author = {Carrizosa, Emilio and Ramırez-Ayerbe, Jasone and Romero, Dolores},
+ title = {Generating {{Collective Counterfactual Explanations}} in {{Score-Based Classification}} via {{Mathematical Optimization}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Article{cascarino2022explainable,
+ author = {Cascarino, Giuseppe and Moscatelli, Mirko and Parlapiano, Fabio},
+ title = {Explainable {{Artificial Intelligence}}: Interpreting Default Forecasting Models Based on {{Machine Learning}}},
+ number = {674},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Bank of Italy Occasional Paper},
+ year = {2022},
+}
+
+@Article{chandola2009anomaly,
+ author = {Chandola, Varun and Banerjee, Arindam and Kumar, Vipin},
+ title = {Anomaly Detection: {{A}} Survey},
+ number = {3},
+ pages = {1--58},
+ volume = {41},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {ACM computing surveys (CSUR)},
+ year = {2009},
+}
+
+@Article{chapelle2011empirical,
+ author = {Chapelle, Olivier and Li, Lihong},
+ title = {An Empirical Evaluation of Thompson Sampling},
+ pages = {2249--2257},
+ volume = {24},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in neural information processing systems},
+ year = {2011},
+}
+
+@Article{chetty2011adjustment,
+ author = {Chetty, Raj and Friedman, John N and Olsen, Tore and Pistaferri, Luigi},
+ title = {Adjustment Costs, Firm Responses, and Micro vs. Macro Labor Supply Elasticities: {{Evidence}} from {{Danish}} Tax Records},
+ number = {2},
+ pages = {749--804},
+ volume = {126},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The quarterly journal of economics},
+ year = {2011},
+}
+
+@Article{cortes1995supportvector,
+ author = {Cortes, Corinna and Vapnik, Vladimir},
+ title = {Support-Vector Networks},
+ number = {3},
+ pages = {273--297},
+ volume = {20},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Machine learning},
+ year = {1995},
+}
+
+@Article{crawford2019variable,
+ author = {Crawford, Lorin and Flaxman, Seth R and Runcie, Daniel E and West, Mike},
+ title = {Variable Prioritization in Nonlinear Black Box Methods: {{A}} Genetic Association Case Study},
+ number = {2},
+ pages = {958},
+ volume = {13},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The annals of applied statistics},
+ year = {2019},
+}
+
+@InProceedings{dai2022counterfactual,
+ author = {Dai, Xinyue and Keane, Mark T and Shalloo, Laurence and Ruelle, Elodie and Byrne, Ruth MJ},
+ title = {Counterfactual Explanations for Prediction and Diagnosis in Xai},
+ eventtitle = {Proceedings of the 2022 {{AAAI}}/{{ACM Conference}} on {{AI}}, {{Ethics}}, and {{Society}}},
+ pages = {215--226},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2022},
+}
+
+@Article{danielsson2021artificial,
+ author = {Danielsson, Jon and Macrae, Robert and Uthemann, Andreas},
+ title = {Artificial Intelligence and Systemic Risk},
+ pages = {106290},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Banking \& Finance},
+ year = {2021},
+}
+
+@Article{daxberger2021laplace,
+ author = {Daxberger, Erik and Kristiadi, Agustinus and Immer, Alexander and Eschenhagen, Runa and Bauer, Matthias and Hennig, Philipp},
+ title = {Laplace {{Redux-Effortless Bayesian Deep Learning}}},
+ volume = {34},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in Neural Information Processing Systems},
+ year = {2021},
+}
+
+@Article{dehejia1999causal,
+ author = {Dehejia, Rajeev H and Wahba, Sadek},
+ title = {Causal Effects in Nonexperimental Studies: {{Reevaluating}} the Evaluation of Training Programs},
+ number = {448},
+ pages = {1053--1062},
+ volume = {94},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of the American statistical Association},
+ year = {1999},
+}
+
+@Article{dell2010persistent,
+ author = {Dell, Melissa},
+ title = {The Persistent Effects of {{Peru}}'s Mining Mita},
+ number = {6},
+ pages = {1863--1903},
+ volume = {78},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica : journal of the Econometric Society},
+ shortjournal = {Econometrica},
+ year = {2010},
+}
+
+@Article{denhengst2020reinforcement,
+ author = {den Hengst, Floris and Grua, Eoin Martino and el Hassouni, Ali and Hoogendoorn, Mark},
+ title = {Reinforcement Learning for Personalization: {{A}} Systematic Literature Review},
+ issue = {Preprint},
+ pages = {1--41},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Data Science},
+ options = {useprefix=true},
+ year = {2020},
+}
+
+@Article{deoliveira2021framework,
+ author = {de Oliveira, Raphael Mazzine Barbosa and Martens, David},
+ title = {A Framework and Benchmarking Study for Counterfactual Generating Methods on Tabular Data},
+ number = {16},
+ pages = {7274},
+ volume = {11},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Applied Sciences},
+ options = {useprefix=true},
+ year = {2021},
+}
+
+@Article{dhurandhar2018explanations,
+ author = {Dhurandhar, Amit and Chen, Pin-Yu and Luss, Ronny and Tu, Chun-Chen and Ting, Paishun and Shanmugam, Karthikeyan and Das, Payel},
+ title = {Explanations Based on the Missing: {{Towards}} Contrastive Explanations with Pertinent Negatives},
+ volume = {31},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in neural information processing systems},
+ year = {2018},
+}
+
+@InProceedings{dombrowski2021diffeomorphic,
+ author = {Dombrowski, Ann-Kathrin and Gerken, Jan E and Kessel, Pan},
+ booktitle = {{{ICML Workshop}} on {{Invertible Neural Networks}}, {{Normalizing Flows}}, and {{Explicit Likelihood Models}}},
+ title = {Diffeomorphic Explanations with Normalizing Flows},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@InProceedings{dorffner1996neural,
+ author = {Dorffner, Georg},
+ booktitle = {Neural Network World},
+ title = {Neural Networks for Time Series Processing},
+ publisher = {{Citeseer}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1996},
+}
+
+@Article{epstein1979stability,
+ author = {Epstein, Seymour},
+ title = {The Stability of Behavior: {{I}}. {{On}} Predicting Most of the People Much of the Time.},
+ number = {7},
+ pages = {1097},
+ volume = {37},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of personality and social psychology},
+ year = {1979},
+}
+
+@Online{barocas2022fairness,
+ author = {Solon Barocas and Moritz Hardt and Arvind Narayanan},
+ title = {Fairness and Machine Learning},
+ url = {https://fairmlbook.org/index.html},
+ urldate = {2022-11-08},
+ bdsk-url-1 = {https://fairmlbook.org/index.html},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ month = dec,
+ year = {2022},
+}
+
+@Article{falk2006clean,
+ author = {Falk, Armin and Ichino, Andrea},
+ title = {Clean Evidence on Peer Effects},
+ number = {1},
+ pages = {39--57},
+ volume = {24},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of labor economics},
+ year = {2006},
+}
+
+@Unpublished{fan2020interpretability,
+ author = {Fan, Fenglei and Xiong, Jinjun and Wang, Ge},
+ title = {On Interpretability of Artificial Neural Networks},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2001.02522},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Article{fang2011dynamic,
+ author = {Fang, Hanming and Gavazza, Alessandro},
+ title = {Dynamic Inefficiencies in an Employment-Based Health Insurance System: {{Theory}} and Evidence},
+ number = {7},
+ pages = {3047--77},
+ volume = {101},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {American Economic Review},
+ year = {2011},
+}
+
+@Article{fehr2000cooperation,
+ author = {Fehr, Ernst and Gachter, Simon},
+ title = {Cooperation and Punishment in Public Goods Experiments},
+ number = {4},
+ pages = {980--994},
+ volume = {90},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {American Economic Review},
+ year = {2000},
+}
+
+@Article{fix1951important,
+ author = {Fix, E and Hodges, J},
+ title = {An Important Contribution to Nonparametric Discriminant Analysis and Density Estimation},
+ number = {57},
+ pages = {233--238},
+ volume = {3},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {International Statistical Review},
+ year = {1951},
+}
+
+@Book{friedman2008monetary,
+ author = {Friedman, Milton and Schwartz, Anna Jacobson},
+ title = {A Monetary History of the {{United States}}, 1867-1960},
+ publisher = {{Princeton University Press}},
+ volume = {14},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2008},
+}
+
+@InProceedings{gal2016dropout,
+ author = {Gal, Yarin and Ghahramani, Zoubin},
+ booktitle = {International Conference on Machine Learning},
+ title = {Dropout as a Bayesian Approximation: {{Representing}} Model Uncertainty in Deep Learning},
+ pages = {1050--1059},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2016},
+}
+
+@InProceedings{gal2017deep,
+ author = {Gal, Yarin and Islam, Riashat and Ghahramani, Zoubin},
+ booktitle = {International {{Conference}} on {{Machine Learning}}},
+ title = {Deep Bayesian Active Learning with Image Data},
+ pages = {1183--1192},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2017},
+}
+
+@Article{galizzi2019external,
+ author = {Galizzi, Matteo M and Navarro-Martinez, Daniel},
+ title = {On the External Validity of Social Preference Games: A Systematic Lab-Field Study},
+ number = {3},
+ pages = {976--1002},
+ volume = {65},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Management Science},
+ year = {2019},
+}
+
+@Article{gama2014survey,
+ author = {Gama, Jo{\~a}o and {\v Z}liobait{\.e}, Indr{\.e} and Bifet, Albert and Pechenizkiy, Mykola and Bouchachia, Abdelhamid},
+ title = {A Survey on Concept Drift Adaptation},
+ number = {4},
+ pages = {1--37},
+ volume = {46},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {ACM computing surveys (CSUR)},
+ year = {2014},
+}
+
+@Unpublished{garivier2008upperconfidence,
+ author = {Garivier, Aur{\'e}lien and Moulines, Eric},
+ title = {On Upper-Confidence Bound Policies for Non-Stationary Bandit Problems},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {0805.3415},
+ eprinttype = {arxiv},
+ year = {2008},
+}
+
+@Book{gelman2013bayesian,
+ author = {Gelman, Andrew and Carlin, John B and Stern, Hal S and Dunson, David B and Vehtari, Aki and Rubin, Donald B},
+ title = {Bayesian Data Analysis},
+ publisher = {{CRC press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2013},
+}
+
+@Article{gilbert1998immune,
+ author = {Gilbert, Daniel T and Pinel, Elizabeth C and Wilson, Timothy D and Blumberg, Stephen J and Wheatley, Thalia P},
+ title = {Immune Neglect: A Source of Durability Bias in Affective Forecasting.},
+ number = {3},
+ pages = {617},
+ volume = {75},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of personality and social psychology},
+ year = {1998},
+}
+
+@Article{gneezy2006uncertainty,
+ author = {Gneezy, Uri and List, John A and Wu, George},
+ title = {The Uncertainty Effect: {{When}} a Risky Prospect Is Valued Less than Its Worst Possible Outcome},
+ number = {4},
+ pages = {1283--1309},
+ volume = {121},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Quarterly Journal of Economics},
+ year = {2006},
+}
+
+@InCollection{goan2020bayesian,
+ author = {Goan, Ethan and Fookes, Clinton},
+ booktitle = {Case {{Studies}} in {{Applied Bayesian Data Science}}},
+ title = {Bayesian {{Neural Networks}}: {{An Introduction}} and {{Survey}}},
+ pages = {45--87},
+ publisher = {{Springer}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Article{goldsmith-pinkham2013social,
+ author = {Goldsmith-Pinkham, Paul and Imbens, Guido W},
+ title = {Social Networks and the Identification of Peer Effects},
+ number = {3},
+ pages = {253--264},
+ volume = {31},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Business \& Economic Statistics},
+ year = {2013},
+}
+
+@Unpublished{goodfellow2014explaining,
+ author = {Goodfellow, Ian J and Shlens, Jonathon and Szegedy, Christian},
+ title = {Explaining and Harnessing Adversarial Examples},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1412.6572},
+ eprinttype = {arxiv},
+ year = {2014},
+}
+
+@Book{goodfellow2016deep,
+ author = {Goodfellow, Ian and Bengio, Yoshua and Courville, Aaron},
+ title = {Deep {{Learning}}},
+ publisher = {{MIT Press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2016},
+}
+
+@Article{goodfriend2005incredible,
+ author = {Goodfriend, Marvin and King, Robert G},
+ title = {The Incredible {{Volcker}} Disinflation},
+ number = {5},
+ pages = {981--1015},
+ volume = {52},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Monetary Economics},
+ year = {2005},
+}
+
+@Article{graham2017econometric,
+ author = {Graham, Bryan S},
+ title = {An Econometric Model of Network Formation with Degree Heterogeneity},
+ number = {4},
+ pages = {1033--1063},
+ volume = {85},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica : journal of the Econometric Society},
+ shortjournal = {Econometrica},
+ year = {2017},
+}
+
+@Article{greene2012econometric,
+ author = {Greene, William H},
+ title = {Econometric Analysis, 71e},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Stern School of Business, New York University},
+ year = {2012},
+}
+
+@Article{grether1979economic,
+ author = {Grether, David M and Plott, Charles R},
+ title = {Economic Theory of Choice and the Preference Reversal Phenomenon},
+ number = {4},
+ pages = {623--638},
+ volume = {69},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The American Economic Review},
+ year = {1979},
+}
+
+@Article{gretton2012kernel,
+ author = {Gretton, Arthur and Borgwardt, Karsten M and Rasch, Malte J and Sch{\"o}lkopf, Bernhard and Smola, Alexander},
+ title = {A Kernel Two-Sample Test},
+ number = {1},
+ pages = {723--773},
+ volume = {13},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Journal of Machine Learning Research},
+ year = {2012},
+}
+
+@Unpublished{griffith2020name,
+ author = {Griffith, Alan},
+ title = {Name {{Your Friends}}, but {{Only Five}}? {{The Importance}} of {{Censoring}} in {{Peer Effects Estimates}} Using {{Social Network Data}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Unpublished{grinsztajn2022why,
+ author = {Grinsztajn, L{\'e}o and Oyallon, Edouard and Varoquaux, Ga{\"e}l},
+ title = {Why Do Tree-Based Models Still Outperform Deep Learning on Tabular Data?},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2207.08815},
+ eprinttype = {arxiv},
+ year = {2022},
+}
+
+@Misc{group2020detailed,
+ author = {Group, Open COVID-19 Data Working},
+ title = {Detailed {{Epidemiological Data}} from the {{COVID-19 Outbreak}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@InProceedings{gupta2011thompson,
+ author = {Gupta, Neha and Granmo, Ole-Christoffer and Agrawala, Ashok},
+ booktitle = {2011 10th {{International Conference}} on {{Machine Learning}} and {{Applications}} and {{Workshops}}},
+ title = {Thompson Sampling for Dynamic Multi-Armed Bandits},
+ pages = {484--489},
+ publisher = {{IEEE}},
+ volume = {1},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2011},
+}
+
+@Book{hamilton2020time,
+ author = {Hamilton, James Douglas},
+ title = {Time Series Analysis},
+ publisher = {{Princeton university press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Article{hamon2020robustness,
+ author = {Hamon, Ronan and Junklewitz, Henrik and Sanchez, Ignacio},
+ title = {Robustness and Explainability of Artificial Intelligence},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Publications Office of the European Union},
+ year = {2020},
+}
+
+@Article{hamzacebi2008improving,
+ author = {Hamza{\c c}ebi, Co{\c s}kun},
+ title = {Improving Artificial Neural Networks' Performance in Seasonal Time Series Forecasting},
+ number = {23},
+ pages = {4550--4559},
+ volume = {178},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Information Sciences},
+ year = {2008},
+}
+
+@InProceedings{hanneke2007bound,
+ author = {Hanneke, Steve},
+ booktitle = {Proceedings of the 24th International Conference on {{Machine}} Learning},
+ title = {A Bound on the Label Complexity of Agnostic Active Learning},
+ pages = {353--360},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2007},
+}
+
+@Article{hansen2020virtue,
+ author = {Hansen, Kristian Bondo},
+ title = {The Virtue of Simplicity: {{On}} Machine Learning Models in Algorithmic Trading},
+ number = {1},
+ pages = {2053951720926558},
+ volume = {7},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Big Data \& Society},
+ year = {2020},
+}
+
+@Article{hartland2006multiarmed,
+ author = {Hartland, C{\'e}dric and Gelly, Sylvain and Baskiotis, Nicolas and Teytaud, Olivier and Sebag, Michele},
+ title = {Multi-Armed Bandit, Dynamic Environments and Meta-Bandits},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2006},
+}
+
+@Article{heckman1985alternative,
+ author = {Heckman, James J and Robb Jr, Richard},
+ title = {Alternative Methods for Evaluating the Impact of Interventions: {{An}} Overview},
+ number = {1-2},
+ pages = {239--267},
+ volume = {30},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of econometrics},
+ year = {1985},
+}
+
+@Article{hershfield2011increasing,
+ author = {Hershfield, Hal E and Goldstein, Daniel G and Sharpe, William F and Fox, Jesse and Yeykelis, Leo and Carstensen, Laura L and Bailenson, Jeremy N},
+ title = {Increasing Saving Behavior through Age-Progressed Renderings of the Future Self},
+ issue = {SPL},
+ pages = {S23--S37},
+ volume = {48},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Marketing Research},
+ year = {2011},
+}
+
+@InProceedings{ho1995random,
+ author = {Ho, Tin Kam},
+ booktitle = {Proceedings of 3rd International Conference on Document Analysis and Recognition},
+ title = {Random Decision Forests},
+ pages = {278--282},
+ publisher = {{IEEE}},
+ volume = {1},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1995},
+}
+
+@Article{hochreiter1997long,
+ author = {Hochreiter, Sepp and Schmidhuber, J{\"u}rgen},
+ title = {Long Short-Term Memory},
+ number = {8},
+ pages = {1735--1780},
+ volume = {9},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Neural computation},
+ year = {1997},
+}
+
+@Unpublished{hoff2021bayesoptimal,
+ author = {Hoff, Peter},
+ title = {Bayes-Optimal Prediction with Frequentist Coverage Control},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2105.14045},
+ eprinttype = {arxiv},
+ file = {:/Users/FA31DU/Zotero/storage/IQK27WVA/Hoff - 2021 - Bayes-optimal prediction with frequentist coverage.pdf:;:/Users/FA31DU/Zotero/storage/K8EAZA25/2105.html:},
+ year = {2021},
+}
+
+@Misc{hoffman1994german,
+ author = {Hoffman, Hans},
+ title = {German {{Credit Data}}},
+ url = {https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)},
+ bdsk-url-1 = {https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1994},
+}
+
+@Online{hoffmanGermanCreditData1994,
+ author = {Hoffman, Hans},
+ title = {German {{Credit Data}}},
+ url = {https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)},
+ bdsk-url-1 = {https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1994},
+}
+
+@Unpublished{houlsby2011bayesian,
+ author = {Houlsby, Neil and Husz{\'a}r, Ferenc and Ghahramani, Zoubin and Lengyel, M{\'a}t{\'e}},
+ title = {Bayesian Active Learning for Classification and Preference Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1112.5745},
+ eprinttype = {arxiv},
+ year = {2011},
+}
+
+@Article{hsee1996evaluability,
+ author = {Hsee, Christopher K},
+ title = {The Evaluability Hypothesis: {{An}} Explanation for Preference Reversals between Joint and Separate Evaluations of Alternatives},
+ number = {3},
+ pages = {247--257},
+ volume = {67},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Organizational behavior and human decision processes},
+ year = {1996},
+}
+
+@Article{hsee2004music,
+ author = {Hsee, Christopher K and Rottenstreich, Yuval},
+ title = {Music, Pandas, and Muggers: On the Affective Psychology of Value.},
+ number = {1},
+ pages = {23},
+ volume = {133},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Experimental Psychology: General},
+ year = {2004},
+}
+
+@Article{hsieh2016social,
+ author = {Hsieh, Chih-Sheng and Lee, Lung Fei},
+ title = {A Social Interactions Model with Endogenous Friendship Formation and Selectivity},
+ number = {2},
+ pages = {301--319},
+ volume = {31},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Applied Econometrics},
+ year = {2016},
+}
+
+@Unpublished{immer2020improving,
+ author = {Immer, Alexander and Korzepa, Maciej and Bauer, Matthias},
+ title = {Improving Predictions of Bayesian Neural Networks via Local Linearization},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2008.08400},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Unpublished{innes2018fashionable,
+ author = {Innes, Michael and Saba, Elliot and Fischer, Keno and Gandhi, Dhairya and Rudilosso, Marco Concetto and Joy, Neethu Mariya and Karmali, Tejan and Pal, Avik and Shah, Viral},
+ title = {Fashionable Modelling with Flux},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1811.01457},
+ eprinttype = {arxiv},
+ year = {2018},
+}
+
+@Article{innes2018flux,
+ author = {Innes, Mike},
+ title = {Flux: {{Elegant}} Machine Learning with {{Julia}}},
+ number = {25},
+ pages = {602},
+ volume = {3},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Open Source Software},
+ year = {2018},
+}
+
+@Unpublished{ish-horowicz2019interpreting,
+ author = {Ish-Horowicz, Jonathan and Udwin, Dana and Flaxman, Seth and Filippi, Sarah and Crawford, Lorin},
+ title = {Interpreting Deep Neural Networks through Variable Importance},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1901.09839},
+ eprinttype = {arxiv},
+ year = {2019},
+}
+
+@InProceedings{jabbari2017fairness,
+ author = {Jabbari, Shahin and Joseph, Matthew and Kearns, Michael and Morgenstern, Jamie and Roth, Aaron},
+ booktitle = {International {{Conference}} on {{Machine Learning}}},
+ title = {Fairness in Reinforcement Learning},
+ pages = {1617--1626},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2017},
+}
+
+@Article{jackson2007meeting,
+ author = {Jackson, Matthew O and Rogers, Brian W},
+ title = {Meeting Strangers and Friends of Friends: {{How}} Random Are Social Networks?},
+ number = {3},
+ pages = {890--915},
+ volume = {97},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {American Economic Review},
+ year = {2007},
+}
+
+@Unpublished{jeanneret2022diffusion,
+ author = {Jeanneret, Guillaume and Simon, Lo{\"\i}c and Jurie, Fr{\'e}d{\'e}ric},
+ title = {Diffusion {{Models}} for {{Counterfactual Explanations}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2203.15636},
+ eprinttype = {arxiv},
+ year = {2022},
+}
+
+@Article{johansson2005failure,
+ author = {Johansson, Petter and Hall, Lars and Sikstr{\"o}m, Sverker and Olsson, Andreas},
+ title = {Failure to Detect Mismatches between Intention and Outcome in a Simple Decision Task},
+ number = {5745},
+ pages = {116--119},
+ volume = {310},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Science (New York, N.Y.)},
+ shortjournal = {Science},
+ year = {2005},
+}
+
+@Article{johnsson2021estimation,
+ author = {Johnsson, Ida and Moon, Hyungsik Roger},
+ title = {Estimation of Peer Effects in Endogenous Social Networks: {{Control}} Function Approach},
+ number = {2},
+ pages = {328--345},
+ volume = {103},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Review of Economics and Statistics},
+ year = {2021},
+}
+
+@Article{jolliffe2003modified,
+ author = {Jolliffe, Ian T and Trendafilov, Nickolay T and Uddin, Mudassir},
+ title = {A Modified Principal Component Technique Based on the {{LASSO}}},
+ number = {3},
+ pages = {531--547},
+ volume = {12},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of computational and Graphical Statistics},
+ year = {2003},
+}
+
+@Article{joseph2021forecasting,
+ author = {Joseph, Andreas and Kalamara, Eleni and Kapetanios, George and Potjagailo, Galina},
+ title = {Forecasting Uk Inflation Bottom Up},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Unpublished{joshi2019realistic,
+ author = {Joshi, Shalmali and Koyejo, Oluwasanmi and Vijitbenjaronk, Warut and Kim, Been and Ghosh, Joydeep},
+ title = {Towards Realistic Individual Recourse and Actionable Explanations in Black-Box Decision Making Systems},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1907.09615},
+ eprinttype = {arxiv},
+ year = {2019},
+}
+
+@Unpublished{jospin2020handson,
+ author = {Jospin, Laurent Valentin and Buntine, Wray and Boussaid, Farid and Laga, Hamid and Bennamoun, Mohammed},
+ title = {Hands-on {{Bayesian Neural Networks}}--a {{Tutorial}} for {{Deep Learning Users}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2007.06823},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Misc{kaggle2011give,
+ author = {Kaggle},
+ title = {Give Me Some Credit, {{Improve}} on the State of the Art in Credit Scoring by Predicting the Probability That Somebody Will Experience Financial Distress in the next Two Years.},
+ url = {https://www.kaggle.com/c/GiveMeSomeCredit},
+ bdsk-url-1 = {https://www.kaggle.com/c/GiveMeSomeCredit},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ publisher = {{Kaggle}},
+ year = {2011},
+}
+
+@online{kagglecompetitionGiveMeCredit,
+ author = {Kaggle Competition},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ title = {Give Me Some Credit, {{Improve}} on the State of the Art in Credit Scoring by Predicting the Probability That Somebody Will Experience Financial Distress in the next Two Years.},
+ url = {https://www.kaggle.com/c/GiveMeSomeCredit},
+ bdsk-url-1 = {https://www.kaggle.com/c/GiveMeSomeCredit}}
+
+@Article{kahneman1979prospect,
+ author = {Kahneman, Daniel and Tversky, Amos},
+ title = {Prospect {{Theory}}: {{An Analysis}} of {{Decision}} under {{Risk}}},
+ pages = {263--291},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica: Journal of the Econometric Society},
+ year = {1979},
+}
+
+@Article{kahneman1990experimental,
+ author = {Kahneman, Daniel and Knetsch, Jack L and Thaler, Richard H},
+ title = {Experimental Tests of the Endowment Effect and the {{Coase}} Theorem},
+ number = {6},
+ pages = {1325--1348},
+ volume = {98},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of political Economy},
+ year = {1990},
+}
+
+@Article{kahneman1992reference,
+ author = {Kahneman, Daniel},
+ title = {Reference Points, Anchors, Norms, and Mixed Feelings},
+ number = {2},
+ pages = {296--312},
+ volume = {51},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Organizational behavior and human decision processes},
+ year = {1992},
+}
+
+@Unpublished{karimi2020algorithmic,
+ author = {Karimi, Amir-Hossein and Von K{\"u}gelgen, Julius and Sch{\"o}lkopf, Bernhard and Valera, Isabel},
+ title = {Algorithmic Recourse under Imperfect Causal Knowledge: A Probabilistic Approach},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2006.06831},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Unpublished{karimi2020survey,
+ author = {Karimi, Amir-Hossein and Barthe, Gilles and Sch{\"o}lkopf, Bernhard and Valera, Isabel},
+ title = {A Survey of Algorithmic Recourse: Definitions, Formulations, Solutions, and Prospects},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2010.04050},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@InProceedings{karimi2021algorithmic,
+ author = {Karimi, Amir-Hossein and Sch{\"o}lkopf, Bernhard and Valera, Isabel},
+ booktitle = {Proceedings of the 2021 {{ACM Conference}} on {{Fairness}}, {{Accountability}}, and {{Transparency}}},
+ title = {Algorithmic Recourse: From Counterfactual Explanations to Interventions},
+ pages = {353--362},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@InProceedings{kaur2020interpreting,
+ author = {Kaur, Harmanpreet and Nori, Harsha and Jenkins, Samuel and Caruana, Rich and Wallach, Hanna and Wortman Vaughan, Jennifer},
+ booktitle = {Proceedings of the 2020 {{CHI}} Conference on Human Factors in Computing Systems},
+ title = {Interpreting Interpretability: Understanding Data Scientists' Use of Interpretability Tools for Machine Learning},
+ pages = {1--14},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Article{kehoe2021defence,
+ author = {Kehoe, Aidan and Wittek, Peter and Xue, Yanbo and Pozas-Kerstjens, Alejandro},
+ title = {Defence against Adversarial Attacks Using Classical and Quantum-Enhanced {{Boltzmann}} Machines},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Machine Learning: Science and Technology},
+ year = {2021},
+}
+
+@Unpublished{kendall2017what,
+ author = {Kendall, Alex and Gal, Yarin},
+ title = {What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1703.04977},
+ eprinttype = {arxiv},
+ year = {2017},
+}
+
+@Article{kihoro2004seasonal,
+ author = {Kihoro, J and Otieno, RO and Wafula, C},
+ title = {Seasonal Time Series Forecasting: {{A}} Comparative Study of {{ARIMA}} and {{ANN}} Models},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2004},
+}
+
+@Book{kilian2017structural,
+ author = {Kilian, Lutz and L{\"u}tkepohl, Helmut},
+ title = {Structural Vector Autoregressive Analysis},
+ publisher = {{Cambridge University Press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2017},
+}
+
+@Unpublished{kingma2014adam,
+ author = {Kingma, Diederik P and Ba, Jimmy},
+ title = {Adam: {{A}} Method for Stochastic Optimization},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1412.6980},
+ eprinttype = {arxiv},
+ year = {2014},
+}
+
+@Article{kirsch2019batchbald,
+ author = {Kirsch, Andreas and Van Amersfoort, Joost and Gal, Yarin},
+ title = {Batchbald: {{Efficient}} and Diverse Batch Acquisition for Deep Bayesian Active Learning},
+ pages = {7026--7037},
+ volume = {32},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in neural information processing systems},
+ year = {2019},
+}
+
+@Unpublished{kuiper2021exploring,
+ author = {Kuiper, Ouren and van den Berg, Martin and van den Burgt, Joost and Leijnen, Stefan},
+ title = {Exploring {{Explainable AI}} in the {{Financial Sector}}: {{Perspectives}} of {{Banks}} and {{Supervisory Authorities}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2111.02244},
+ eprinttype = {arxiv},
+ year = {2021},
+}
+
+@Article{kydland1982time,
+ author = {Kydland, Finn E and Prescott, Edward C},
+ title = {Time to Build and Aggregate Fluctuations},
+ pages = {1345--1370},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica: Journal of the Econometric Society},
+ year = {1982},
+}
+
+@Unpublished{lachapelle2019gradientbased,
+ author = {Lachapelle, S{\'e}bastien and Brouillard, Philippe and Deleu, Tristan and Lacoste-Julien, Simon},
+ title = {Gradient-Based Neural Dag Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1906.02226},
+ eprinttype = {arxiv},
+ year = {2019},
+}
+
+@InProceedings{lakkaraju2020how,
+ author = {Lakkaraju, Himabindu and Bastani, Osbert},
+ booktitle = {Proceedings of the {{AAAI}}/{{ACM Conference}} on {{AI}}, {{Ethics}}, and {{Society}}},
+ title = {" {{How}} Do {{I}} Fool You?" {{Manipulating User Trust}} via {{Misleading Black Box Explanations}}},
+ pages = {79--85},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@InProceedings{lakkaraju2020how,
+ author = {Lakkaraju, Himabindu and Bastani, Osbert},
+ booktitle = {Proceedings of the {{AAAI}}/{{ACM Conference}} on {{AI}}, {{Ethics}}, and {{Society}}},
+ title = {" {{How Do I Fool You}}?" {{Manipulating User Trust}} via {{Misleading Black Box Explanations}}},
+ pages = {79--85},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Unpublished{lakshminarayanan2016simple,
+ author = {Lakshminarayanan, Balaji and Pritzel, Alexander and Blundell, Charles},
+ title = {Simple and Scalable Predictive Uncertainty Estimation Using Deep Ensembles},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1612.01474},
+ eprinttype = {arxiv},
+ year = {2016},
+}
+
+@Unpublished{laugel2017inverse,
+ author = {Laugel, Thibault and Lesot, Marie-Jeanne and Marsala, Christophe and Renard, Xavier and Detyniecki, Marcin},
+ title = {Inverse Classification for Comparison-Based Interpretability in Machine Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1712.08443},
+ eprinttype = {arxiv},
+ shortjournal = {arXiv preprint arXiv:1712.08443},
+ year = {2017},
+}
+
+@Thesis{lawrence2001variational,
+ author = {Lawrence, Neil David},
+ title = {Variational Inference in Probabilistic Models},
+ type = {phdthesis},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ school = {{University of Cambridge}},
+ year = {2001},
+}
+
+@Article{lecun1998mnist,
+ author = {LeCun, Yann},
+ title = {The {{MNIST}} Database of Handwritten Digits},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ shortjournal = {http://yann. lecun. com/exdb/mnist/},
+ year = {1998},
+}
+
+@Article{lee2003best,
+ author = {Lee, Lung-fei},
+ title = {Best Spatial Two-Stage Least Squares Estimators for a Spatial Autoregressive Model with Autoregressive Disturbances},
+ number = {4},
+ pages = {307--335},
+ volume = {22},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometric Reviews},
+ year = {2003},
+}
+
+@Article{lerner2013financial,
+ author = {Lerner, Jennifer S and Li, Ye and Weber, Elke U},
+ title = {The Financial Costs of Sadness},
+ number = {1},
+ pages = {72--79},
+ volume = {24},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Psychological science},
+ year = {2013},
+}
+
+@Article{list2004neoclassical,
+ author = {List, John A},
+ title = {Neoclassical Theory versus Prospect Theory: {{Evidence}} from the Marketplace},
+ number = {2},
+ pages = {615--625},
+ volume = {72},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Econometrica : journal of the Econometric Society},
+ shortjournal = {Econometrica},
+ year = {2004},
+}
+
+@Article{lucas1976econometric,
+ author = {Lucas, JR},
+ title = {Econometric Policy Evaluation: A Critique `, in {{K}}. {{Brunner}} and {{A Meltzer}}, {{The Phillips}} Curve and Labor Markets, {{North Holland}}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {1976},
+}
+
+@InProceedings{lundberg2017unified,
+ author = {Lundberg, Scott M and Lee, Su-In},
+ booktitle = {Proceedings of the 31st International Conference on Neural Information Processing Systems},
+ title = {A Unified Approach to Interpreting Model Predictions},
+ pages = {4768--4777},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2017},
+}
+
+@Book{lutkepohl2005new,
+ author = {L{\"u}tkepohl, Helmut},
+ title = {New Introduction to Multiple Time Series Analysis},
+ publisher = {{Springer Science \& Business Media}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2005},
+}
+
+@Article{madrian2001power,
+ author = {Madrian, Brigitte C and Shea, Dennis F},
+ title = {The Power of Suggestion: {{Inertia}} in 401 (k) Participation and Savings Behavior},
+ number = {4},
+ pages = {1149--1187},
+ volume = {116},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Quarterly journal of economics},
+ year = {2001},
+}
+
+@Book{manning2008introduction,
+ author = {Manning, Christopher D and Sch{\"u}tze, Hinrich and Raghavan, Prabhakar},
+ title = {Introduction to Information Retrieval},
+ publisher = {{Cambridge university press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2008},
+}
+
+@misc{manokhin2022awesome,
+ author = {Manokhin, Valery},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ title = {Awesome Conformal Prediction}}
+
+@Article{manski1993identification,
+ author = {Manski, Charles F},
+ title = {Identification of Endogenous Social Effects: {{The}} Reflection Problem},
+ number = {3},
+ pages = {531--542},
+ volume = {60},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The review of economic studies},
+ year = {1993},
+}
+
+@Article{markle2018goals,
+ author = {Markle, Alex and Wu, George and White, Rebecca and Sackett, Aaron},
+ title = {Goals as Reference Points in Marathon Running: {{A}} Novel Test of Reference Dependence},
+ number = {1},
+ pages = {19--50},
+ volume = {56},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Risk and Uncertainty},
+ year = {2018},
+}
+
+@Article{masini2021machine,
+ author = {Masini, Ricardo P and Medeiros, Marcelo C and Mendes, Eduardo F},
+ title = {Machine Learning Advances for Time Series Forecasting},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Economic Surveys},
+ year = {2021},
+}
+
+@Article{mccracken2016fredmd,
+ author = {McCracken, Michael W and Ng, Serena},
+ title = {{{FRED-MD}}: {{A}} Monthly Database for Macroeconomic Research},
+ number = {4},
+ pages = {574--589},
+ volume = {34},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Business \& Economic Statistics},
+ year = {2016},
+}
+
+@Article{mcculloch1990logical,
+ author = {McCulloch, Warren S and Pitts, Walter},
+ title = {A Logical Calculus of the Ideas Immanent in Nervous Activity},
+ number = {1},
+ pages = {99--115},
+ volume = {52},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Bulletin of mathematical biology},
+ year = {1990},
+}
+
+@Article{migut2015visualizing,
+ author = {Migut, MA and Worring, Marcel and Veenman, Cor J},
+ title = {Visualizing Multi-Dimensional Decision Boundaries in {{2D}}},
+ number = {1},
+ pages = {273--295},
+ volume = {29},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Data Mining and Knowledge Discovery},
+ year = {2015},
+}
+
+@Article{miller2019explanation,
+ author = {Miller, Tim},
+ title = {Explanation in Artificial Intelligence: {{Insights}} from the Social Sciences},
+ pages = {1--38},
+ volume = {267},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Artificial intelligence},
+ year = {2019},
+}
+
+@InProceedings{miller2020strategic,
+ author = {Miller, John and Milli, Smitha and Hardt, Moritz},
+ booktitle = {Proceedings of the 37th {{International Conference}} on {{Machine Learning}}},
+ title = {Strategic {{Classification}} Is {{Causal Modeling}} in {{Disguise}}},
+ eventtitle = {International {{Conference}} on {{Machine Learning}}},
+ pages = {6917--6926},
+ publisher = {{PMLR}},
+ url = {https://proceedings.mlr.press/v119/miller20b.html},
+ urldate = {2022-11-03},
+ abstract = {Consequential decision-making incentivizes individuals to strategically adapt their behavior to the specifics of the decision rule. While a long line of work has viewed strategic adaptation as gaming and attempted to mitigate its effects, recent work has instead sought to design classifiers that incentivize individuals to improve a desired quality. Key to both accounts is a cost function that dictates which adaptations are rational to undertake. In this work, we develop a causal framework for strategic adaptation. Our causal perspective clearly distinguishes between gaming and improvement and reveals an important obstacle to incentive design. We prove any procedure for designing classifiers that incentivize improvement must inevitably solve a non-trivial causal inference problem. We show a similar result holds for designing cost functions that satisfy the requirements of previous work. With the benefit of hindsight, our results show much of the prior work on strategic classification is causal modeling in disguise.},
+ bdsk-url-1 = {https://proceedings.mlr.press/v119/miller20b.html},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ file = {:/Users/FA31DU/Zotero/storage/46I2QMPI/Miller et al. - 2020 - Strategic Classification is Causal Modeling in Dis.pdf:;:/Users/FA31DU/Zotero/storage/NWREET6B/Miller et al. - 2020 - Strategic Classification is Causal Modeling in Dis.pdf:},
+ issn = {2640-3498},
+ langid = {english},
+ month = nov,
+ year = {2020},
+}
+
+@Article{mischel1988nature,
+ author = {Mischel, Walter and Shoda, Yuichi and Peake, Philip K},
+ title = {The Nature of Adolescent Competencies Predicted by Preschool Delay of Gratification.},
+ number = {4},
+ pages = {687},
+ volume = {54},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of personality and social psychology},
+ year = {1988},
+}
+
+@InProceedings{mittelstadt2019explaining,
+ author = {Mittelstadt, Brent and Russell, Chris and Wachter, Sandra},
+ booktitle = {Proceedings of the Conference on Fairness, Accountability, and Transparency},
+ title = {Explaining Explanations in {{AI}}},
+ pages = {279--288},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2019},
+}
+
+@Book{molnar2020interpretable,
+ author = {Molnar, Christoph},
+ title = {Interpretable Machine Learning},
+ publisher = {{Lulu. com}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Book{morgan2015counterfactuals,
+ author = {Morgan, Stephen L and Winship, Christopher},
+ title = {Counterfactuals and Causal Inference},
+ publisher = {{Cambridge University Press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2015},
+}
+
+@Article{mosteller1951experimental,
+ author = {Mosteller, Frederick and Nogee, Philip},
+ title = {An Experimental Measurement of Utility},
+ number = {5},
+ pages = {371--404},
+ volume = {59},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Political Economy},
+ year = {1951},
+}
+
+@InProceedings{mothilal2020explaining,
+ author = {Mothilal, Ramaravind K and Sharma, Amit and Tan, Chenhao},
+ booktitle = {Proceedings of the 2020 {{Conference}} on {{Fairness}}, {{Accountability}}, and {{Transparency}}},
+ title = {Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations},
+ pages = {607--617},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Book{murphy2012machine,
+ author = {Murphy, Kevin P},
+ title = {Machine Learning: A Probabilistic Perspective},
+ publisher = {{MIT press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2012},
+}
+
+@Book{murphy2012machine,
+ author = {Murphy, Kevin P},
+ title = {Machine Learning: {{A}} Probabilistic Perspective},
+ publisher = {{MIT press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2012},
+}
+
+@Book{murphy2022probabilistic,
+ author = {Murphy, Kevin P},
+ title = {Probabilistic {{Machine Learning}}: {{An}} Introduction},
+ publisher = {{MIT Press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2022},
+}
+
+@Article{nagel1995unraveling,
+ author = {Nagel, Rosemarie},
+ title = {Unraveling in Guessing Games: {{An}} Experimental Study},
+ number = {5},
+ pages = {1313--1326},
+ volume = {85},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The American Economic Review},
+ year = {1995},
+}
+
+@Unpublished{navarro-martinez2021bridging,
+ author = {Navarro-Martinez, Daniel and Wang, Xinghua},
+ title = {Bridging the Gap between the Lab and the Field: {{Dictator}} Games and Donations},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@InProceedings{nelson2015evaluating,
+ author = {Nelson, Kevin and Corbin, George and Anania, Mark and Kovacs, Matthew and Tobias, Jeremy and Blowers, Misty},
+ booktitle = {2015 {{IEEE Symposium}} on {{Computational Intelligence}} for {{Security}} and {{Defense Applications}} ({{CISDA}})},
+ title = {Evaluating Model Drift in Machine Learning Algorithms},
+ pages = {1--8},
+ publisher = {{IEEE}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2015},
+}
+
+@Book{nocedal2006numerical,
+ author = {Nocedal, Jorge and Wright, Stephen},
+ title = {Numerical Optimization},
+ publisher = {{Springer Science \& Business Media}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2006},
+}
+
+@Misc{oecd2021artificial,
+ author = {{OECD}},
+ title = {Artificial {{Intelligence}}, {{Machine Learning}} and {{Big Data}} in {{Finance}}: {{Opportunities}}, {{Challenges}} and {{Implications}} for {{Policy Makers}}},
+ url = {https://www.oecd.org/finance/financial-markets/Artificial-intelligence-machine-learning-big-data-in-finance.pdf},
+ bdsk-url-1 = {https://www.oecd.org/finance/financial-markets/Artificial-intelligence-machine-learning-big-data-in-finance.pdf},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Online{oecdArtificialIntelligenceMachine2021,
+ author = {{OECD}},
+ title = {Artificial {{Intelligence}}, {{Machine Learning}} and {{Big Data}} in {{Finance}}: {{Opportunities}}, {{Challenges}} and {{Implications}} for {{Policy Makers}}},
+ url = {https://www.oecd.org/finance/financial-markets/Artificial-intelligence-machine-learning-big-data-in-finance.pdf},
+ bdsk-url-1 = {https://www.oecd.org/finance/financial-markets/Artificial-intelligence-machine-learning-big-data-in-finance.pdf},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ publisher = {{OECD}},
+ year = {2021},
+}
+
+@Book{oneil2016weapons,
+ author = {O'Neil, Cathy},
+ title = {Weapons of Math Destruction: {{How}} Big Data Increases Inequality and Threatens Democracy},
+ publisher = {{Crown}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2016},
+}
+
+@Article{pace1997sparse,
+ author = {Pace, R Kelley and Barry, Ronald},
+ title = {Sparse Spatial Autoregressions},
+ number = {3},
+ pages = {291--297},
+ volume = {33},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Statistics \& Probability Letters},
+ year = {1997},
+}
+
+@Unpublished{parr2018matrix,
+ author = {Parr, Terence and Howard, Jeremy},
+ title = {The Matrix Calculus You Need for Deep Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1802.01528},
+ eprinttype = {arxiv},
+ year = {2018},
+}
+
+@Unpublished{pawelczyk2021carla,
+ author = {Pawelczyk, Martin and Bielawski, Sascha and van den Heuvel, Johannes and Richter, Tobias and Kasneci, Gjergji},
+ title = {Carla: A Python Library to Benchmark Algorithmic Recourse and Counterfactual Explanation Algorithms},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2108.00783},
+ eprinttype = {arxiv},
+ year = {2021},
+}
+
+@Book{pearl2018book,
+ author = {Pearl, Judea and Mackenzie, Dana},
+ title = {The Book of Why: The New Science of Cause and Effect},
+ publisher = {{Basic books}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2018},
+}
+
+@Article{pearl2019seven,
+ author = {Pearl, Judea},
+ title = {The Seven Tools of Causal Inference, with Reflections on Machine Learning},
+ number = {3},
+ pages = {54--60},
+ volume = {62},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Communications of the ACM},
+ year = {2019},
+}
+
+@Article{pedregosa2011scikitlearn,
+ author = {Pedregosa, Fabian and Varoquaux, Ga{\"e}l and Gramfort, Alexandre and Michel, Vincent and Thirion, Bertrand and Grisel, Olivier and Blondel, Mathieu and Prettenhofer, Peter and Weiss, Ron and Dubourg, Vincent and others},
+ title = {Scikit-Learn: {{Machine}} Learning in {{Python}}},
+ pages = {2825--2830},
+ volume = {12},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {the Journal of machine Learning research},
+ year = {2011},
+}
+
+@Book{perry2010economic,
+ author = {Perry, George L and Tobin, James},
+ title = {Economic {{Events}}, {{Ideas}}, and {{Policies}}: The 1960s and After},
+ publisher = {{Brookings Institution Press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2010},
+}
+
+@Article{pfaff2008var,
+ author = {Pfaff, Bernhard and others},
+ title = {{{VAR}}, {{SVAR}} and {{SVEC}} Models: {{Implementation}} within {{R}} Package Vars},
+ number = {4},
+ pages = {1--32},
+ volume = {27},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Statistical Software},
+ year = {2008},
+}
+
+@Book{pindyck2014microeconomics,
+ author = {Pindyck, Robert S and Rubinfeld, Daniel L},
+ title = {Microeconomics},
+ publisher = {{Pearson Education}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2014},
+}
+
+@Article{pope2011numbers,
+ author = {Pope, Devin and Simonsohn, Uri},
+ title = {Round Numbers as Goals: {{Evidence}} from Baseball, {{SAT}} Takers, and the Lab},
+ number = {1},
+ pages = {71--79},
+ volume = {22},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Psychological science},
+ year = {2011},
+}
+
+@InProceedings{poyiadzi2020face,
+ author = {Poyiadzi, Rafael and Sokol, Kacper and Santos-Rodriguez, Raul and De Bie, Tijl and Flach, Peter},
+ booktitle = {Proceedings of the {{AAAI}}/{{ACM Conference}} on {{AI}}, {{Ethics}}, and {{Society}}},
+ title = {{{FACE}}: {{Feasible}} and Actionable Counterfactual Explanations},
+ pages = {344--350},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Article{qu2015estimating,
+ author = {Qu, Xi and Lee, Lung-fei},
+ title = {Estimating a Spatial Autoregressive Model with an Endogenous Spatial Weight Matrix},
+ number = {2},
+ pages = {209--232},
+ volume = {184},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of Econometrics},
+ year = {2015},
+}
+
+@Article{rabanser2019failing,
+ author = {Rabanser, Stephan and G{\"u}nnemann, Stephan and Lipton, Zachary},
+ title = {Failing Loudly: {{An}} Empirical Study of Methods for Detecting Dataset Shift},
+ volume = {32},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in Neural Information Processing Systems},
+ year = {2019},
+}
+
+@Unpublished{raghunathan2019adversarial,
+ author = {Raghunathan, Aditi and Xie, Sang Michael and Yang, Fanny and Duchi, John C and Liang, Percy},
+ title = {Adversarial Training Can Hurt Generalization},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1906.06032},
+ eprinttype = {arxiv},
+ year = {2019},
+}
+
+@Unpublished{raj2017taming,
+ author = {Raj, Vishnu and Kalyani, Sheetal},
+ title = {Taming Non-Stationary Bandits: {{A Bayesian}} Approach},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1707.09727},
+ eprinttype = {arxiv},
+ year = {2017},
+}
+
+@InProceedings{rasmussen2003gaussian,
+ author = {Rasmussen, Carl Edward},
+ booktitle = {Summer School on Machine Learning},
+ title = {Gaussian Processes in Machine Learning},
+ pages = {63--71},
+ publisher = {{Springer}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2003},
+}
+
+@InProceedings{ribeiro2016why,
+ author = {Ribeiro, Marco Tulio and Singh, Sameer and Guestrin, Carlos},
+ booktitle = {Proceedings of the 22nd {{ACM SIGKDD}} International Conference on Knowledge Discovery and Data Mining},
+ title = {"{{Why}} Should i Trust You?" {{Explaining}} the Predictions of Any Classifier},
+ pages = {1135--1144},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2016},
+}
+
+@Article{romer1989does,
+ author = {Romer, Christina D and Romer, David H},
+ title = {Does Monetary Policy Matter? {{A}} New Test in the Spirit of {{Friedman}} and {{Schwartz}}},
+ pages = {121--170},
+ volume = {4},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {NBER macroeconomics annual},
+ year = {1989},
+}
+
+@Article{rudin2019stop,
+ author = {Rudin, Cynthia},
+ title = {Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead},
+ number = {5},
+ pages = {206--215},
+ volume = {1},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Nature Machine Intelligence},
+ year = {2019},
+}
+
+@Article{sacerdote2001peer,
+ author = {Sacerdote, Bruce},
+ title = {Peer Effects with Random Assignment: {{Results}} for {{Dartmouth}} Roommates},
+ number = {2},
+ pages = {681--704},
+ volume = {116},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The Quarterly journal of economics},
+ year = {2001},
+}
+
+@Article{sadinle2019least,
+ author = {Sadinle, Mauricio and Lei, Jing and Wasserman, Larry},
+ title = {Least Ambiguous Set-Valued Classifiers with Bounded Error Levels},
+ number = {525},
+ pages = {223--234},
+ volume = {114},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ file = {:/Users/FA31DU/Zotero/storage/YXQ8N76A/Sadinle et al. - 2019 - Least ambiguous set-valued classifiers with bounde.pdf:;:/Users/FA31DU/Zotero/storage/ZHB56F3V/01621459.2017.html:},
+ journal = {Journal of the American Statistical Association},
+ publisher = {{Taylor \& Francis}},
+ year = {2019},
+}
+
+@InProceedings{satopaa2011finding,
+ author = {Satopaa, Ville and Albrecht, Jeannie and Irwin, David and Raghavan, Barath},
+ booktitle = {2011 31st International Conference on Distributed Computing Systems Workshops},
+ title = {Finding a" Kneedle" in a Haystack: {{Detecting}} Knee Points in System Behavior},
+ pages = {166--171},
+ publisher = {{IEEE}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2011},
+}
+
+@InProceedings{schut2021generating,
+ author = {Schut, Lisa and Key, Oscar and Mc Grath, Rory and Costabello, Luca and Sacaleanu, Bogdan and Gal, Yarin and others},
+ booktitle = {International {{Conference}} on {{Artificial Intelligence}} and {{Statistics}}},
+ title = {Generating {{Interpretable Counterfactual Explanations By Implicit Minimisation}} of {{Epistemic}} and {{Aleatoric Uncertainties}}},
+ pages = {1756--1764},
+ publisher = {{PMLR}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2021},
+}
+
+@Book{schutze2008introduction,
+ author = {Sch{\"u}tze, Hinrich and Manning, Christopher D and Raghavan, Prabhakar},
+ title = {Introduction to Information Retrieval},
+ publisher = {{Cambridge University Press Cambridge}},
+ volume = {39},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2008},
+}
+
+@Article{shafir1993reasonbased,
+ author = {Shafir, Eldar and Simonson, Itamar and Tversky, Amos},
+ title = {Reason-Based Choice},
+ number = {1-2},
+ pages = {11--36},
+ volume = {49},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Cognition},
+ year = {1993},
+}
+
+@Article{simonson1989choice,
+ author = {Simonson, Itamar},
+ title = {Choice Based on Reasons: {{The}} Case of Attraction and Compromise Effects},
+ number = {2},
+ pages = {158--174},
+ volume = {16},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of consumer research},
+ year = {1989},
+}
+
+@Article{sims1986are,
+ author = {Sims, Christopher A and others},
+ title = {Are Forecasting Models Usable for Policy Analysis?},
+ issue = {Win},
+ pages = {2--16},
+ volume = {10},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Quarterly Review},
+ year = {1986},
+}
+
+@InProceedings{slack2020fooling,
+ author = {Slack, Dylan and Hilgard, Sophie and Jia, Emily and Singh, Sameer and Lakkaraju, Himabindu},
+ booktitle = {Proceedings of the {{AAAI}}/{{ACM Conference}} on {{AI}}, {{Ethics}}, and {{Society}}},
+ title = {Fooling Lime and Shap: {{Adversarial}} Attacks on Post Hoc Explanation Methods},
+ pages = {180--186},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2020},
+}
+
+@Article{slack2021counterfactual,
+ author = {Slack, Dylan and Hilgard, Anna and Lakkaraju, Himabindu and Singh, Sameer},
+ title = {Counterfactual Explanations Can Be Manipulated},
+ volume = {34},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Advances in Neural Information Processing Systems},
+ year = {2021},
+}
+
+@Article{slovic1974who,
+ author = {Slovic, Paul and Tversky, Amos},
+ title = {Who Accepts {{Savage}}'s Axiom?},
+ number = {6},
+ pages = {368--373},
+ volume = {19},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Behavioral science},
+ year = {1974},
+}
+
+@Unpublished{spooner2021counterfactual,
+ author = {Spooner, Thomas and Dervovic, Danial and Long, Jason and Shepard, Jon and Chen, Jiahao and Magazzeni, Daniele},
+ title = {Counterfactual {{Explanations}} for {{Arbitrary Regression Models}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2106.15212},
+ eprinttype = {arxiv},
+ shortjournal = {arXiv preprint arXiv:2106.15212},
+ year = {2021},
+}
+
+@Article{srivastava2014dropout,
+ author = {Srivastava, Nitish and Hinton, Geoffrey and Krizhevsky, Alex and Sutskever, Ilya and Salakhutdinov, Ruslan},
+ title = {Dropout: A Simple Way to Prevent Neural Networks from Overfitting},
+ number = {1},
+ pages = {1929--1958},
+ volume = {15},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The journal of machine learning research},
+ year = {2014},
+}
+
+@Unpublished{stanton2022bayesian,
+ author = {Stanton, Samuel and Maddox, Wesley and Wilson, Andrew Gordon},
+ title = {Bayesian {{Optimization}} with {{Conformal Coverage Guarantees}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2210.12496},
+ eprinttype = {arxiv},
+ file = {:/Users/FA31DU/Zotero/storage/XFGZAB9J/Stanton et al. - 2022 - Bayesian Optimization with Conformal Coverage Guar.pdf:;:/Users/FA31DU/Zotero/storage/RPWYDPVW/2210.html:},
+ year = {2022},
+}
+
+@Article{sturm2014simple,
+ author = {Sturm, Bob L},
+ title = {A Simple Method to Determine If a Music Information Retrieval System Is a ``Horse''},
+ number = {6},
+ pages = {1636--1644},
+ volume = {16},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {IEEE Transactions on Multimedia},
+ year = {2014},
+}
+
+@Article{sunstein2003libertarian,
+ author = {Sunstein, Cass R and Thaler, Richard H},
+ title = {Libertarian Paternalism Is Not an Oxymoron},
+ pages = {1159--1202},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {The University of Chicago Law Review},
+ year = {2003},
+}
+
+@Book{sutton2018reinforcement,
+ author = {Sutton, Richard S and Barto, Andrew G},
+ title = {Reinforcement Learning: {{An}} Introduction},
+ publisher = {{MIT press}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2018},
+}
+
+@Unpublished{szegedy2013intriguing,
+ author = {Szegedy, Christian and Zaremba, Wojciech and Sutskever, Ilya and Bruna, Joan and Erhan, Dumitru and Goodfellow, Ian and Fergus, Rob},
+ title = {Intriguing Properties of Neural Networks},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1312.6199},
+ eprinttype = {arxiv},
+ year = {2013},
+}
+
+@Article{thaler1981empirical,
+ author = {Thaler, Richard},
+ title = {Some Empirical Evidence on Dynamic Inconsistency},
+ number = {3},
+ pages = {201--207},
+ volume = {8},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Economics letters},
+ year = {1981},
+}
+
+@Article{thaler2004more,
+ author = {Thaler, Richard H and Benartzi, Shlomo},
+ title = {Save More Tomorrow{\texttrademark}: {{Using}} Behavioral Economics to Increase Employee Saving},
+ number = {S1},
+ pages = {S164--S187},
+ volume = {112},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of political Economy},
+ year = {2004},
+}
+
+@Article{tversky1981framing,
+ author = {Tversky, Amos and Kahneman, Daniel},
+ title = {The Framing of Decisions and the Psychology of Choice},
+ number = {4481},
+ pages = {453--458},
+ volume = {211},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Science (New York, N.Y.)},
+ shortjournal = {science},
+ year = {1981},
+}
+
+@Article{ungemach2011how,
+ author = {Ungemach, Christoph and Stewart, Neil and Reimers, Stian},
+ title = {How Incidental Values from the Environment Affect Decisions about Money, Risk, and Delay},
+ number = {2},
+ pages = {253--260},
+ volume = {22},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Psychological Science},
+ year = {2011},
+}
+
+@Unpublished{upadhyay2021robust,
+ author = {Upadhyay, Sohini and Joshi, Shalmali and Lakkaraju, Himabindu},
+ title = {Towards {{Robust}} and {{Reliable Algorithmic Recourse}}},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2102.13620},
+ eprinttype = {arxiv},
+ year = {2021},
+}
+
+@InProceedings{ustun2019actionable,
+ author = {Ustun, Berk and Spangher, Alexander and Liu, Yang},
+ booktitle = {Proceedings of the {{Conference}} on {{Fairness}}, {{Accountability}}, and {{Transparency}}},
+ title = {Actionable Recourse in Linear Classification},
+ pages = {10--19},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2019},
+}
+
+@Article{vanboven2000egocentric,
+ author = {Van Boven, Leaf and Dunning, David and Loewenstein, George},
+ title = {Egocentric Empathy Gaps between Owners and Buyers: Misperceptions of the Endowment Effect.},
+ number = {1},
+ pages = {66},
+ volume = {79},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of personality and social psychology},
+ year = {2000},
+}
+
+@Book{varshney2022trustworthy,
+ author = {Varshney, Kush R.},
+ title = {Trustworthy {{Machine Learning}}},
+ publisher = {{Independently Published}},
+ address = {{Chappaqua, NY, USA}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2022},
+}
+
+@Unpublished{verma2020counterfactual,
+ author = {Verma, Sahil and Dickerson, John and Hines, Keegan},
+ title = {Counterfactual Explanations for Machine Learning: {{A}} Review},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2010.10596},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Article{verstyuk2020modeling,
+ author = {Verstyuk, Sergiy},
+ title = {Modeling Multivariate Time Series in Economics: {{From}} Auto-Regressions to Recurrent Neural Networks},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Available at SSRN 3589337},
+ year = {2020},
+}
+
+@Article{wachter2017counterfactual,
+ author = {Wachter, Sandra and Mittelstadt, Brent and Russell, Chris},
+ title = {Counterfactual Explanations without Opening the Black Box: {{Automated}} Decisions and the {{GDPR}}},
+ pages = {841},
+ volume = {31},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Harv. JL \& Tech.},
+ year = {2017},
+}
+
+@Article{wang2018optimal,
+ author = {Wang, HaiYing and Zhu, Rong and Ma, Ping},
+ title = {Optimal Subsampling for Large Sample Logistic Regression},
+ number = {522},
+ pages = {829--844},
+ volume = {113},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Journal of the American Statistical Association},
+ year = {2018},
+}
+
+@Book{wasserman2006all,
+ author = {Wasserman, Larry},
+ title = {All of Nonparametric Statistics},
+ publisher = {{Springer Science \& Business Media}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2006},
+}
+
+@Book{wasserman2013all,
+ author = {Wasserman, Larry},
+ title = {All of Statistics: A Concise Course in Statistical Inference},
+ publisher = {{Springer Science \& Business Media}},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ year = {2013},
+}
+
+@Article{widmer1996learning,
+ author = {Widmer, Gerhard and Kubat, Miroslav},
+ title = {Learning in the Presence of Concept Drift and Hidden Contexts},
+ number = {1},
+ pages = {69--101},
+ volume = {23},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Machine learning},
+ year = {1996},
+}
+
+@Unpublished{wilson2020case,
+ author = {Wilson, Andrew Gordon},
+ title = {The Case for {{Bayesian}} Deep Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {2001.10995},
+ eprinttype = {arxiv},
+ year = {2020},
+}
+
+@Article{witten2009penalized,
+ author = {Witten, Daniela M and Tibshirani, Robert and Hastie, Trevor},
+ title = {A Penalized Matrix Decomposition, with Applications to Sparse Principal Components and Canonical Correlation Analysis},
+ number = {3},
+ pages = {515--534},
+ volume = {10},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Biostatistics (Oxford, England)},
+ shortjournal = {Biostatistics},
+ year = {2009},
+}
+
+@Article{xu2020epidemiological,
+ author = {Xu, Bo and Gutierrez, Bernardo and Mekaru, Sumiko and Sewalk, Kara and Goodwin, Lauren and Loskill, Alyssa and Cohn, Emily and Hswen, Yulin and Hill, Sarah C. and Cobo, Maria M and Zarebski, Alexander and Li, Sabrina and Wu, Chieh-Hsi and Hulland, Erin and Morgan, Julia and Wang, Lin and O'Brien, Katelynn and Scarpino, Samuel V. and Brownstein, John S. and Pybus, Oliver G. and Pigott, David M. and Kraemer, Moritz U. G.},
+ title = {Epidemiological Data from the {{COVID-19}} Outbreak, Real-Time Case Information},
+ doi = {doi.org/10.1038/s41597-020-0448-0},
+ number = {106},
+ volume = {7},
+ bdsk-url-1 = {https://doi.org/10.1038/s41597-020-0448-0},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Scientific Data},
+ year = {2020},
+}
+
+@Article{yeh2009comparisons,
+ author = {Yeh, I-Cheng and Lien, Che-hui},
+ title = {The Comparisons of Data Mining Techniques for the Predictive Accuracy of Probability of Default of Credit Card Clients},
+ number = {2},
+ pages = {2473--2480},
+ volume = {36},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Expert systems with applications},
+ year = {2009},
+}
+
+@Article{zhang1998forecasting,
+ author = {Zhang, Guoqiang and Patuwo, B Eddy and Hu, Michael Y},
+ title = {Forecasting with Artificial Neural Networks:: {{The}} State of the Art},
+ number = {1},
+ pages = {35--62},
+ volume = {14},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {International journal of forecasting},
+ year = {1998},
+}
+
+@Article{zhang2003time,
+ author = {Zhang, G Peter},
+ title = {Time Series Forecasting Using a Hybrid {{ARIMA}} and Neural Network Model},
+ pages = {159--175},
+ volume = {50},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {Neurocomputing},
+ year = {2003},
+}
+
+@Unpublished{zheng2018dags,
+ author = {Zheng, Xun and Aragam, Bryon and Ravikumar, Pradeep and Xing, Eric P},
+ title = {Dags with No Tears: {{Continuous}} Optimization for Structure Learning},
+ archiveprefix = {arXiv},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ eprint = {1803.01422},
+ eprinttype = {arxiv},
+ year = {2018},
+}
+
+@Article{zhu2015optimal,
+ author = {Zhu, Rong and Ma, Ping and Mahoney, Michael W and Yu, Bin},
+ title = {Optimal Subsampling Approaches for Large Sample Linear Regression},
+ pages = {arXiv--1509},
+ date-added = {2022-12-13 12:58:01 +0100},
+ date-modified = {2022-12-13 12:58:01 +0100},
+ journal = {arXiv},
+ year = {2015},
+}
+
+@Article{barber2021predictive,
+ author = {Barber, Rina Foygel and Candès, Emmanuel J. and Ramdas, Aaditya and Tibshirani, Ryan J.},
+ title = {Predictive inference with the jackknife+},
+ doi = {10.1214/20-AOS1965},
+ issn = {0090-5364, 2168-8966},
+ number = {1},
+ pages = {486--507},
+ urldate = {2022-12-13},
+ volume = {49},
+ abstract = {This paper introduces the jackknife+, which is a novel method for constructing predictive confidence intervals. Whereas the jackknife outputs an interval centered at the predicted response of a test point, with the width of the interval determined by the quantiles of leave-one-out residuals, the jackknife+ also uses the leave-one-out predictions at the test point to account for the variability in the fitted regression function. Assuming exchangeable training samples, we prove that this crucial modification permits rigorous coverage guarantees regardless of the distribution of the data points, for any algorithm that treats the training points symmetrically. Such guarantees are not possible for the original jackknife and we demonstrate examples where the coverage rate may actually vanish. Our theoretical and empirical analysis reveals that the jackknife and the jackknife+ intervals achieve nearly exact coverage and have similar lengths whenever the fitting algorithm obeys some form of stability. Further, we extend the jackknife+ to \$K\$-fold cross validation and similarly establish rigorous coverage properties. Our methods are related to cross-conformal prediction proposed by Vovk (Ann. Math. Artif. Intell. 74 (2015) 9–28) and we discuss connections.},
+ file = {:Barber2021 - Predictive Inference with the Jackknife+.pdf:PDF},
+ journal = {The Annals of Statistics},
+ keywords = {62F40, 62G08, 62G09, conformal inference, cross-validation, distribution-free, jackknife, leave-one-out, stability},
+ month = feb,
+ publisher = {Institute of Mathematical Statistics},
+ year = {2021},
+}
+
+@TechReport{chouldechova2018frontiers,
+ author = {Chouldechova, Alexandra and Roth, Aaron},
+ title = {The {Frontiers} of {Fairness} in {Machine} {Learning}},
+ doi = {10.48550/arXiv.1810.08810},
+ eprint = {1810.08810},
+ note = {arXiv:1810.08810 [cs, stat] type: article},
+ abstract = {The last few years have seen an explosion of academic and popular interest in algorithmic fairness. Despite this interest and the volume and velocity of work that has been produced recently, the fundamental science of fairness in machine learning is still in a nascent state. In March 2018, we convened a group of experts as part of a CCC visioning workshop to assess the state of the field, and distill the most promising research directions going forward. This report summarizes the findings of that workshop. Along the way, it surveys recent theoretical work in the field and points towards promising directions for research.},
+ archiveprefix = {arxiv},
+ file = {:chouldechova2018frontiers - The Frontiers of Fairness in Machine Learning.pdf:PDF},
+ keywords = {Computer Science - Machine Learning, Computer Science - Data Structures and Algorithms, Computer Science - Computer Science and Game Theory, Statistics - Machine Learning},
+ month = oct,
+ school = {arXiv},
+ year = {2018},
+}
+
+@Article{pawelczyk2022probabilistically,
+ author = {Pawelczyk, Martin and Datta, Teresa and van-den-Heuvel, Johannes and Kasneci, Gjergji and Lakkaraju, Himabindu},
+ title = {Probabilistically {Robust} {Recourse}: {Navigating} the {Trade}-offs between {Costs} and {Robustness} in {Algorithmic} {Recourse}},
+ file = {:pawelczyk2022probabilistically - Probabilistically Robust Recourse_ Navigating the Trade Offs between Costs and Robustness in Algorithmic Recourse.pdf:PDF},
+ journal = {arXiv preprint arXiv:2203.06768},
+ shorttitle = {Probabilistically {Robust} {Recourse}},
+ year = {2022},
+}
+
+@InProceedings{stutz2022learning,
+ author = {Stutz, David and Dvijotham, Krishnamurthy Dj and Cemgil, Ali Taylan and Doucet, Arnaud},
+ title = {Learning {Optimal} {Conformal} {Classifiers}},
+ language = {en},
+ url = {https://openreview.net/forum?id=t8O-4LKFVx},
+ urldate = {2023-02-13},
+ abstract = {Modern deep learning based classifiers show very high accuracy on test data but this does not provide sufficient guarantees for safe deployment, especially in high-stake AI applications such as medical diagnosis. Usually, predictions are obtained without a reliable uncertainty estimate or a formal guarantee. Conformal prediction (CP) addresses these issues by using the classifier's predictions, e.g., its probability estimates, to predict confidence sets containing the true class with a user-specified probability. However, using CP as a separate processing step after training prevents the underlying model from adapting to the prediction of confidence sets. Thus, this paper explores strategies to differentiate through CP during training with the goal of training model with the conformal wrapper end-to-end. In our approach, conformal training (ConfTr), we specifically "simulate" conformalization on mini-batches during training. Compared to standard training, ConfTr reduces the average confidence set size (inefficiency) of state-of-the-art CP methods applied after training. Moreover, it allows to "shape" the confidence sets predicted at test time, which is difficult for standard CP. On experiments with several datasets, we show ConfTr can influence how inefficiency is distributed across classes, or guide the composition of confidence sets in terms of the included classes, while retaining the guarantees offered by CP.},
+ file = {:stutz2022learning - Learning Optimal Conformal Classifiers.pdf:PDF},
+ month = may,
+ year = {2022},
+}
+
+@InProceedings{grathwohl2020your,
+ author = {Grathwohl, Will and Wang, Kuan-Chieh and Jacobsen, Joern-Henrik and Duvenaud, David and Norouzi, Mohammad and Swersky, Kevin},
+ title = {Your classifier is secretly an energy based model and you should treat it like one},
+ language = {en},
+ url = {https://openreview.net/forum?id=Hkxzx0NtDB},
+ urldate = {2023-02-13},
+ abstract = {We propose to reinterpret a standard discriminative classifier of p(y{\textbar}x) as an energy based model for the joint distribution p(x, y). In this setting, the standard class probabilities can be easily computed as well as unnormalized values of p(x) and p(x{\textbar}y). Within this framework, standard discriminative architectures may be used and the model can also be trained on unlabeled data. We demonstrate that energy based training of the joint distribution improves calibration, robustness, and out-of-distribution detection while also enabling our models to generate samples rivaling the quality of recent GAN approaches. We improve upon recently proposed techniques for scaling up the training of energy based models and present an approach which adds little overhead compared to standard classification training. Our approach is the first to achieve performance rivaling the state-of-the-art in both generative and discriminative learning within one hybrid model.},
+ file = {:grathwohl2020your - Your Classifier Is Secretly an Energy Based Model and You Should Treat It like One.pdf:PDF},
+ month = mar,
+ year = {2020},
+}
+
+@Book{murphy2023probabilistic,
+ author = {Murphy, Kevin P.},
+ date = {2023},
+ title = {Probabilistic machine learning: {Advanced} topics},
+ publisher = {MIT Press},
+ shorttitle = {Probabilistic machine learning},
+}
+
+@TechReport{artelt2021evaluating,
+ author = {Artelt, André and Vaquet, Valerie and Velioglu, Riza and Hinder, Fabian and Brinkrolf, Johannes and Schilling, Malte and Hammer, Barbara},
+ date = {2021-07},
+ institution = {arXiv},
+ title = {Evaluating {Robustness} of {Counterfactual} {Explanations}},
+ note = {arXiv:2103.02354 [cs] type: article},
+ url = {http://arxiv.org/abs/2103.02354},
+ urldate = {2023-03-24},
+ abstract = {Transparency is a fundamental requirement for decision making systems when these should be deployed in the real world. It is usually achieved by providing explanations of the system's behavior. A prominent and intuitive type of explanations are counterfactual explanations. Counterfactual explanations explain a behavior to the user by proposing actions -- as changes to the input -- that would cause a different (specified) behavior of the system. However, such explanation methods can be unstable with respect to small changes to the input -- i.e. even a small change in the input can lead to huge or arbitrary changes in the output and of the explanation. This could be problematic for counterfactual explanations, as two similar individuals might get very different explanations. Even worse, if the recommended actions differ considerably in their complexity, one would consider such unstable (counterfactual) explanations as individually unfair. In this work, we formally and empirically study the robustness of counterfactual explanations in general, as well as under different models and different kinds of perturbations. Furthermore, we propose that plausible counterfactual explanations can be used instead of closest counterfactual explanations to improve the robustness and consequently the individual fairness of counterfactual explanations.},
+ annotation = {Comment: Rewrite paper to make things more clear; Remove one theorem \& corollary due to buggy proof},
+ file = {:artelt2021evaluating - Evaluating Robustness of Counterfactual Explanations.pdf:PDF},
+ keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence},
+}
+
+@Article{guidotti2022counterfactual,
+ author = {Guidotti, Riccardo},
+ date = {2022-04},
+ journaltitle = {Data Mining and Knowledge Discovery},
+ title = {Counterfactual explanations and how to find them: literature review and benchmarking},
+ doi = {10.1007/s10618-022-00831-6},
+ issn = {1573-756X},
+ language = {en},
+ url = {https://doi.org/10.1007/s10618-022-00831-6},
+ urldate = {2023-03-24},
+ abstract = {Interpretable machine learning aims at unveiling the reasons behind predictions returned by uninterpretable classifiers. One of the most valuable types of explanation consists of counterfactuals. A counterfactual explanation reveals what should have been different in an instance to observe a diverse outcome. For instance, a bank customer asks for a loan that is rejected. The counterfactual explanation consists of what should have been different for the customer in order to have the loan accepted. Recently, there has been an explosion of proposals for counterfactual explainers. The aim of this work is to survey the most recent explainers returning counterfactual explanations. We categorize explainers based on the approach adopted to return the counterfactuals, and we label them according to characteristics of the method and properties of the counterfactuals returned. In addition, we visually compare the explanations, and we report quantitative benchmarking assessing minimality, actionability, stability, diversity, discriminative power, and running time. The results make evident that the current state of the art does not provide a counterfactual explainer able to guarantee all these properties simultaneously.},
+ file = {Full Text PDF:https\://link.springer.com/content/pdf/10.1007%2Fs10618-022-00831-6.pdf:application/pdf},
+ keywords = {Explainable AI, Counterfactual explanations, Contrastive explanations, Interpretable machine learning},
+ shorttitle = {Counterfactual explanations and how to find them},
+}
+
+@TechReport{mahajan2020preserving,
+ author = {Mahajan, Divyat and Tan, Chenhao and Sharma, Amit},
+ date = {2020-06},
+ institution = {arXiv},
+ title = {Preserving {Causal} {Constraints} in {Counterfactual} {Explanations} for {Machine} {Learning} {Classifiers}},
+ doi = {10.48550/arXiv.1912.03277},
+ note = {arXiv:1912.03277 [cs, stat] type: article},
+ url = {http://arxiv.org/abs/1912.03277},
+ urldate = {2023-03-24},
+ abstract = {To construct interpretable explanations that are consistent with the original ML model, counterfactual examples---showing how the model's output changes with small perturbations to the input---have been proposed. This paper extends the work in counterfactual explanations by addressing the challenge of feasibility of such examples. For explanations of ML models in critical domains such as healthcare and finance, counterfactual examples are useful for an end-user only to the extent that perturbation of feature inputs is feasible in the real world. We formulate the problem of feasibility as preserving causal relationships among input features and present a method that uses (partial) structural causal models to generate actionable counterfactuals. When feasibility constraints cannot be easily expressed, we consider an alternative mechanism where people can label generated CF examples on feasibility: whether it is feasible to intervene and realize the candidate CF example from the original input. To learn from this labelled feasibility data, we propose a modified variational auto encoder loss for generating CF examples that optimizes for feasibility as people interact with its output. Our experiments on Bayesian networks and the widely used ''Adult-Income'' dataset show that our proposed methods can generate counterfactual explanations that better satisfy feasibility constraints than existing methods.. Code repository can be accessed here: {\textbackslash}textit\{https://github.com/divyat09/cf-feasibility\}},
+ annotation = {Comment: 2019 NeurIPS Workshop on Do the right thing: Machine learning and Causal Inference for improved decision making},
+ file = {:mahajan2020preserving - Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers.pdf:PDF},
+ keywords = {Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Statistics - Machine Learning},
+}
+
+@TechReport{antoran2023sampling,
+ author = {Antorán, Javier and Padhy, Shreyas and Barbano, Riccardo and Nalisnick, Eric and Janz, David and Hernández-Lobato, José Miguel},
+ date = {2023-03},
+ institution = {arXiv},
+ title = {Sampling-based inference for large linear models, with application to linearised {Laplace}},
+ note = {arXiv:2210.04994 [cs, stat] type: article},
+ url = {http://arxiv.org/abs/2210.04994},
+ urldate = {2023-03-25},
+ abstract = {Large-scale linear models are ubiquitous throughout machine learning, with contemporary application as surrogate models for neural network uncertainty quantification; that is, the linearised Laplace method. Alas, the computational cost associated with Bayesian linear models constrains this method's application to small networks, small output spaces and small datasets. We address this limitation by introducing a scalable sample-based Bayesian inference method for conjugate Gaussian multi-output linear models, together with a matching method for hyperparameter (regularisation) selection. Furthermore, we use a classic feature normalisation method (the g-prior) to resolve a previously highlighted pathology of the linearised Laplace method. Together, these contributions allow us to perform linearised neural network inference with ResNet-18 on CIFAR100 (11M parameters, 100 outputs x 50k datapoints), with ResNet-50 on Imagenet (50M parameters, 1000 outputs x 1.2M datapoints) and with a U-Net on a high-resolution tomographic reconstruction task (2M parameters, 251k output{\textasciitilde}dimensions).},
+ annotation = {Comment: Published at ICLR 2023. This latest Arxiv version is extended with a demonstration of the proposed methods on the Imagenet dataset},
+ file = {arXiv Fulltext PDF:https\://arxiv.org/pdf/2210.04994.pdf:application/pdf},
+ keywords = {Statistics - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Machine Learning},
+}
+
+@Misc{altmeyer2022conformal,
+ author = {Altmeyer, Patrick},
+ date = {2022-10},
+ title = {{Conformal} {Prediction} in {Julia}},
+ language = {en},
+ url = {https://www.paltmeyer.com/blog/posts/conformal-prediction/},
+ urldate = {2023-03-27},
+ abstract = {A (very) gentle introduction to Conformal Prediction in Julia using my new package ConformalPrediction.jl.},
+}
+
+@InProceedings{welling2011bayesian,
+ author = {Welling, M. and Teh, Y.},
+ date = {2011-06},
+ title = {Bayesian {Learning} via {Stochastic} {Gradient} {Langevin} {Dynamics}},
+ url = {https://www.semanticscholar.org/paper/Bayesian-Learning-via-Stochastic-Gradient-Langevin-Welling-Teh/aeed631d6a84100b5e9a021ec1914095c66de415},
+ urldate = {2023-05-15},
+ abstract = {In this paper we propose a new framework for learning from large scale datasets based on iterative learning from small mini-batches. By adding the right amount of noise to a standard stochastic gradient optimization algorithm we show that the iterates will converge to samples from the true posterior distribution as we anneal the stepsize. This seamless transition between optimization and Bayesian posterior sampling provides an inbuilt protection against overfitting. We also propose a practical method for Monte Carlo estimates of posterior statistics which monitors a "sampling threshold" and collects samples after it has been surpassed. We apply the method to three models: a mixture of Gaussians, logistic regression and ICA with natural gradients.},
+ annotation = {[TLDR] This paper proposes a new framework for learning from large scale datasets based on iterative learning from small mini-batches by adding the right amount of noise to a standard stochastic gradient optimization algorithm and shows that the iterates will converge to samples from the true posterior distribution as the authors anneal the stepsize.},
+ file = {:welling_bayesian_2011 - Bayesian Learning Via Stochastic Gradient Langevin Dynamics.html:URL;:welling2011bayesian - Bayesian Learning Via Stochastic Gradient Langevin Dynamics.pdf:PDF},
+}
+
+@Article{gill2010circular,
+ author = {Gill, Jeff and Hangartner, Dominik},
+ date = {2010},
+ journaltitle = {Political Analysis},
+ title = {Circular {Data} in {Political} {Science} and {How} to {Handle} {It}},
+ doi = {10.1093/pan/mpq009},
+ issn = {1047-1987, 1476-4989},
+ language = {en},
+ number = {3},
+ pages = {316--336},
+ url = {https://www.cambridge.org/core/journals/political-analysis/article/circular-data-in-political-science-and-how-to-handle-it/6DF2D9DA60C455E6A48FFB0FF011F747},
+ urldate = {2023-05-15},
+ volume = {18},
+ abstract = {There has been no attention to circular (purely cyclical) data in political science research. We show that such data exist and are mishandled by models that do not take into account the inherently recycling nature of some phenomenon. Clock and calendar effects are the obvious cases, but directional data are observed as well. We describe a standard maximum likelihood regression modeling framework based on the von Mises distribution, then develop a general Bayesian regression procedure for the first time, providing an easy-to-use Metropolis-Hastings sampler for this approach. Applications include a chronographic analysis of U.S. domestic terrorism and directional party preferences in a two-dimensional ideological space for German Bundestag elections. The results demonstrate the importance of circular models to handle periodic and directional data in political science.},
+ file = {Full Text PDF:https\://www.cambridge.org/core/services/aop-cambridge-core/content/view/6DF2D9DA60C455E6A48FFB0FF011F747/S1047198700012493a.pdf/div-class-title-circular-data-in-political-science-and-how-to-handle-it-div.pdf:application/pdf},
+ publisher = {Cambridge University Press},
+}
+
+@InProceedings{liu2023goggle,
+ author = {Liu, Tennison and Qian, Zhaozhi and Berrevoets, Jeroen and Schaar, Mihaela van der},
+ date = {2023-02},
+ title = {{GOGGLE}: {Generative} {Modelling} for {Tabular} {Data} by {Learning} {Relational} {Structure}},
+ language = {en},
+ url = {https://openreview.net/forum?id=fPVRcJqspu},
+ urldate = {2023-05-15},
+ abstract = {Deep generative models learn highly complex and non-linear representations to generate realistic synthetic data. While they have achieved notable success in computer vision and natural language processing, similar advances have been less demonstrable in the tabular domain. This is partially because generative modelling of tabular data entails a particular set of challenges, including heterogeneous relationships, limited number of samples, and difficulties in incorporating prior knowledge. Additionally, unlike their counterparts in image and sequence domain, deep generative models for tabular data almost exclusively employ fully-connected layers, which encode weak inductive biases about relationships between inputs. Real-world data generating processes can often be represented using relational structures, which encode sparse, heterogeneous relationships between variables. In this work, we learn and exploit relational structure underlying tabular data to better model variable dependence, and as a natural means to introduce regularization on relationships and include prior knowledge. Specifically, we introduce GOGGLE, an end-to-end message passing scheme that jointly learns the relational structure and corresponding functional relationships as the basis of generating synthetic samples. Using real-world datasets, we provide empirical evidence that the proposed method is effective in generating realistic synthetic data and exploiting domain knowledge for downstream tasks.},
+ file = {Full Text PDF:https\://openreview.net/pdf?id=fPVRcJqspu:application/pdf},
+ shorttitle = {{GOGGLE}},
+}
+
+@TechReport{du2020implicit,
+ author = {Du, Yilun and Mordatch, Igor},
+ date = {2020-06},
+ institution = {arXiv},
+ title = {Implicit {Generation} and {Generalization} in {Energy}-{Based} {Models}},
+ doi = {10.48550/arXiv.1903.08689},
+ note = {arXiv:1903.08689 [cs, stat] type: article},
+ url = {http://arxiv.org/abs/1903.08689},
+ urldate = {2023-05-16},
+ abstract = {Energy based models (EBMs) are appealing due to their generality and simplicity in likelihood modeling, but have been traditionally difficult to train. We present techniques to scale MCMC based EBM training on continuous neural networks, and we show its success on the high-dimensional data domains of ImageNet32x32, ImageNet128x128, CIFAR-10, and robotic hand trajectories, achieving better samples than other likelihood models and nearing the performance of contemporary GAN approaches, while covering all modes of the data. We highlight some unique capabilities of implicit generation such as compositionality and corrupt image reconstruction and inpainting. Finally, we show that EBMs are useful models across a wide variety of tasks, achieving state-of-the-art out-of-distribution classification, adversarially robust classification, state-of-the-art continual online class learning, and coherent long term predicted trajectory rollouts.},
+ file = {arXiv Fulltext PDF:https\://arxiv.org/pdf/1903.08689.pdf:application/pdf},
+ keywords = {Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Statistics - Machine Learning},
+}
+
+@Comment{jabref-meta: databaseType:biblatex;}
diff --git a/AAAI/contents/table-real-world.tex b/AAAI/contents/table-real-world.tex
new file mode 100644
index 0000000000000000000000000000000000000000..5d67371f126861a32c680147036bde6d875da62f
--- /dev/null
+++ b/AAAI/contents/table-real-world.tex
@@ -0,0 +1,45 @@
+\begin{table}
+
+\caption{Results for real-world datasets: sample averages +/- one standard deviation across counterfactuals. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-real-world} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{llcccc}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{2}{c}{MNIST} & \multicolumn{2}{c}{GMSC} \\
+\cmidrule(l{3pt}r{3pt}){3-4} \cmidrule(l{3pt}r{3pt}){5-6}
+Model & Generator & Unfaithfulness ↓ & Implausibility ↓ & Unfaithfulness ↓ & Implausibility ↓\\
+\midrule
+ & ECCCo & \textbf{19.28 ± 5.01}** & 314.76 ± 32.36*\hphantom{*} & \textbf{79.16 ± 11.67}** & 18.26 ± 4.92**\\
+
+ & REVISE & 188.70 ± 26.18*\hphantom{*} & \textbf{255.26 ± 41.50}** & 186.40 ± 28.06\hphantom{*}\hphantom{*} & \textbf{5.34 ± 2.38}**\\
+
+ & Schut & 211.62 ± 27.13\hphantom{*}\hphantom{*} & 290.56 ± 40.66*\hphantom{*} & 200.98 ± 28.49\hphantom{*}\hphantom{*} & 6.50 ± 2.01**\\
+
+\multirow{-4}{*}{\raggedright\arraybackslash JEM} & Wachter & 222.90 ± 26.56\hphantom{*}\hphantom{*} & 361.88 ± 39.74\hphantom{*}\hphantom{*} & 214.08 ± 45.35\hphantom{*}\hphantom{*} & 61.04 ± 2.58\hphantom{*}\hphantom{*}\\
+\cmidrule{1-6}
+ & ECCCo & \textbf{15.99 ± 3.06}** & 294.72 ± 30.75** & \textbf{83.28 ± 13.26}** & 17.21 ± 4.46**\\
+
+ & REVISE & 173.59 ± 20.65** & \textbf{246.32 ± 37.46}** & 194.24 ± 35.41\hphantom{*}\hphantom{*} & \textbf{4.95 ± 1.26}**\\
+
+ & Schut & 204.36 ± 23.14\hphantom{*}\hphantom{*} & 290.64 ± 39.49*\hphantom{*} & 208.45 ± 34.60\hphantom{*}\hphantom{*} & 6.12 ± 1.91**\\
+
+\multirow{-4}{*}{\raggedright\arraybackslash JEM Ensemble} & Wachter & 217.67 ± 23.78\hphantom{*}\hphantom{*} & 363.23 ± 39.24\hphantom{*}\hphantom{*} & 186.19 ± 33.88\hphantom{*}\hphantom{*} & 60.70 ± 44.32\hphantom{*}\hphantom{*}\\
+\cmidrule{1-6}
+ & ECCCo & \textbf{41.95 ± 6.50}** & 591.58 ± 36.24\hphantom{*}\hphantom{*} & \textbf{75.93 ± 14.27}** & 17.20 ± 3.15**\\
+
+ & REVISE & 365.82 ± 15.35*\hphantom{*} & \textbf{249.49 ± 41.55}** & 196.75 ± 41.25\hphantom{*}\hphantom{*} & \textbf{4.84 ± 0.60}**\\
+
+ & Schut & 379.66 ± 17.16\hphantom{*}\hphantom{*} & 290.07 ± 42.65*\hphantom{*} & 212.00 ± 41.15\hphantom{*}\hphantom{*} & 6.44 ± 1.34**\\
+
+\multirow{-4}{*}{\raggedright\arraybackslash MLP} & Wachter & 386.05 ± 16.60\hphantom{*}\hphantom{*} & 361.83 ± 42.18\hphantom{*}\hphantom{*} & 218.34 ± 53.26\hphantom{*}\hphantom{*} & 45.84 ± 39.39\hphantom{*}\hphantom{*}\\
+\cmidrule{1-6}
+ & ECCCo & \textbf{31.43 ± 3.91}** & 490.88 ± 27.19\hphantom{*}\hphantom{*} & \textbf{73.86 ± 14.63}** & 17.92 ± 4.17**\\
+
+ & REVISE & 337.74 ± 11.89*\hphantom{*} & \textbf{247.67 ± 38.36}** & 207.21 ± 43.20\hphantom{*}\hphantom{*} & \textbf{5.78 ± 2.10}**\\
+
+ & Schut & 354.80 ± 13.05\hphantom{*}\hphantom{*} & 285.79 ± 41.33*\hphantom{*} & 205.36 ± 32.11\hphantom{*}\hphantom{*} & 7.00 ± 2.15**\\
+
+\multirow{-4}{*}{\raggedright\arraybackslash MLP Ensemble} & Wachter & 360.79 ± 14.39\hphantom{*}\hphantom{*} & 357.73 ± 42.55\hphantom{*}\hphantom{*} & 213.71 ± 54.17\hphantom{*}\hphantom{*} & 73.09 ± 64.50\hphantom{*}\hphantom{*}\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table-synthetic.tex b/AAAI/contents/table-synthetic.tex
new file mode 100644
index 0000000000000000000000000000000000000000..df1746fbf5c7717ecbfc890e2f0e7551ab9834d8
--- /dev/null
+++ b/AAAI/contents/table-synthetic.tex
@@ -0,0 +1,37 @@
+\begin{table}
+
+\caption{Results for synthetic datasets: sample averages +/- one standard deviation across counterfactuals. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-synthetic} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{llcccccc}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{2}{c}{Linearly Separable} & \multicolumn{2}{c}{Moons} & \multicolumn{2}{c}{Circles} \\
+\cmidrule(l{3pt}r{3pt}){3-4} \cmidrule(l{3pt}r{3pt}){5-6} \cmidrule(l{3pt}r{3pt}){7-8}
+Model & Generator & Unfaithfulness ↓ & Implausibility ↓ & Unfaithfulness ↓ & Implausibility ↓ & Unfaithfulness ↓ & Implausibility ↓\\
+\midrule
+ & ECCCo & \textbf{0.03 ± 0.06}** & \textbf{0.20 ± 0.08}** & \textbf{0.31 ± 0.30}*\hphantom{*} & \textbf{1.20 ± 0.15}** & 0.52 ± 0.36\hphantom{*}\hphantom{*} & 1.22 ± 0.46\hphantom{*}\hphantom{*}\\
+
+ & ECCCo (no CP) & 0.03 ± 0.06** & 0.20 ± 0.08** & 0.37 ± 0.30*\hphantom{*} & 1.21 ± 0.17** & 0.54 ± 0.39\hphantom{*}\hphantom{*} & 1.21 ± 0.46\hphantom{*}\hphantom{*}\\
+
+ & ECCCo (no EBM) & 0.16 ± 0.11\hphantom{*}\hphantom{*} & 0.34 ± 0.19\hphantom{*}\hphantom{*} & 0.91 ± 0.32\hphantom{*}\hphantom{*} & 1.71 ± 0.25\hphantom{*}\hphantom{*} & 0.70 ± 0.33\hphantom{*}\hphantom{*} & 1.30 ± 0.37\hphantom{*}\hphantom{*}\\
+
+ & REVISE & 0.19 ± 0.03\hphantom{*}\hphantom{*} & 0.41 ± 0.01** & 0.78 ± 0.23\hphantom{*}\hphantom{*} & 1.57 ± 0.26\hphantom{*}\hphantom{*} & \textbf{0.48 ± 0.16}*\hphantom{*} & \textbf{0.95 ± 0.32}*\hphantom{*}\\
+
+ & Schut & 0.39 ± 0.07\hphantom{*}\hphantom{*} & 0.73 ± 0.17\hphantom{*}\hphantom{*} & 0.67 ± 0.27\hphantom{*}\hphantom{*} & 1.50 ± 0.22*\hphantom{*} & 0.54 ± 0.43\hphantom{*}\hphantom{*} & 1.28 ± 0.53\hphantom{*}\hphantom{*}\\
+
+\multirow{-6}{*}{\raggedright\arraybackslash JEM} & Wachter & 0.18 ± 0.10\hphantom{*}\hphantom{*} & 0.44 ± 0.17\hphantom{*}\hphantom{*} & 0.80 ± 0.27\hphantom{*}\hphantom{*} & 1.78 ± 0.24\hphantom{*}\hphantom{*} & 0.68 ± 0.34\hphantom{*}\hphantom{*} & 1.33 ± 0.32\hphantom{*}\hphantom{*}\\
+\cmidrule{1-8}
+ & ECCCo & \textbf{0.29 ± 0.05}** & 0.23 ± 0.06** & 0.80 ± 0.62\hphantom{*}\hphantom{*} & 1.69 ± 0.40\hphantom{*}\hphantom{*} & 0.65 ± 0.53\hphantom{*}\hphantom{*} & 1.17 ± 0.41\hphantom{*}\hphantom{*}\\
+
+ & ECCCo (no CP) & 0.29 ± 0.05** & \textbf{0.23 ± 0.07}** & \textbf{0.79 ± 0.62}\hphantom{*}\hphantom{*} & 1.68 ± 0.42\hphantom{*}\hphantom{*} & \textbf{0.49 ± 0.35}\hphantom{*}\hphantom{*} & 1.19 ± 0.44\hphantom{*}\hphantom{*}\\
+
+ & ECCCo (no EBM) & 0.46 ± 0.05\hphantom{*}\hphantom{*} & 0.28 ± 0.04** & 1.34 ± 0.47\hphantom{*}\hphantom{*} & 1.68 ± 0.47\hphantom{*}\hphantom{*} & 0.84 ± 0.51\hphantom{*}\hphantom{*} & 1.23 ± 0.31\hphantom{*}\hphantom{*}\\
+
+ & REVISE & 0.56 ± 0.05\hphantom{*}\hphantom{*} & 0.41 ± 0.01\hphantom{*}\hphantom{*} & 1.45 ± 0.44\hphantom{*}\hphantom{*} & \textbf{1.64 ± 0.31}\hphantom{*}\hphantom{*} & 0.58 ± 0.52\hphantom{*}\hphantom{*} & \textbf{0.95 ± 0.32}\hphantom{*}\hphantom{*}\\
+
+ & Schut & 0.43 ± 0.06*\hphantom{*} & 0.47 ± 0.36\hphantom{*}\hphantom{*} & 1.45 ± 0.55\hphantom{*}\hphantom{*} & 1.73 ± 0.48\hphantom{*}\hphantom{*} & 0.58 ± 0.37\hphantom{*}\hphantom{*} & 1.23 ± 0.43\hphantom{*}\hphantom{*}\\
+
+\multirow{-6}{*}{\raggedright\arraybackslash MLP} & Wachter & 0.51 ± 0.04\hphantom{*}\hphantom{*} & 0.40 ± 0.08\hphantom{*}\hphantom{*} & 1.32 ± 0.41\hphantom{*}\hphantom{*} & 1.69 ± 0.32\hphantom{*}\hphantom{*} & 0.83 ± 0.50\hphantom{*}\hphantom{*} & 1.24 ± 0.29\hphantom{*}\hphantom{*}\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table_all.tex b/AAAI/contents/table_all.tex
new file mode 100644
index 0000000000000000000000000000000000000000..84a0d04c0153148e937e0490d3dacdd5951389d6
--- /dev/null
+++ b/AAAI/contents/table_all.tex
@@ -0,0 +1,147 @@
+\begin{table}
+
+\caption{All results for all datasets: sample averages +/- one standard deviation over all counterfactuals. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-full} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{ccccccccc}
+\toprule
+Model & Data & Generator & Cost ↓ & Unfaithfulness ↓ & Implausibility ↓ & Redundancy ↑ & Uncertainty ↓ & Validity ↑\\
+\midrule
+ & & ECCCo & 0.74 ± 0.21\hphantom{*}\hphantom{*} & 0.52 ± 0.36\hphantom{*}\hphantom{*} & 1.22 ± 0.46\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & ECCCo (no CP) & 0.72 ± 0.21\hphantom{*}\hphantom{*} & 0.54 ± 0.39\hphantom{*}\hphantom{*} & 1.21 ± 0.46\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & ECCCo (no EBM) & 0.52 ± 0.15\hphantom{*}\hphantom{*} & 0.70 ± 0.33\hphantom{*}\hphantom{*} & 1.30 ± 0.37\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 0.97 ± 0.34\hphantom{*}\hphantom{*} & \textbf{0.48 ± 0.16}*\hphantom{*} & \textbf{0.95 ± 0.32}*\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.50 ± 0.51\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 1.06 ± 0.43\hphantom{*}\hphantom{*} & 0.54 ± 0.43\hphantom{*}\hphantom{*} & 1.28 ± 0.53\hphantom{*}\hphantom{*} & \textbf{0.26 ± 0.25}*\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & \textbf{0.44 ± 0.16}\hphantom{*}\hphantom{*} & 0.68 ± 0.34\hphantom{*}\hphantom{*} & 1.33 ± 0.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.98 ± 0.14\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 0.67 ± 0.19\hphantom{*}\hphantom{*} & 0.65 ± 0.53\hphantom{*}\hphantom{*} & 1.17 ± 0.41\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.09 ± 0.19** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.71 ± 0.16\hphantom{*}\hphantom{*} & \textbf{0.49 ± 0.35}\hphantom{*}\hphantom{*} & 1.19 ± 0.44\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.05 ± 0.16** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.45 ± 0.11\hphantom{*}\hphantom{*} & 0.84 ± 0.51\hphantom{*}\hphantom{*} & 1.23 ± 0.31\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.23*\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 0.96 ± 0.31\hphantom{*}\hphantom{*} & 0.58 ± 0.52\hphantom{*}\hphantom{*} & \textbf{0.95 ± 0.32}\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 0.50 ± 0.51\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 0.57 ± 0.11\hphantom{*}\hphantom{*} & 0.58 ± 0.37\hphantom{*}\hphantom{*} & 1.23 ± 0.43\hphantom{*}\hphantom{*} & \textbf{0.43 ± 0.18}** & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Circles} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & \textbf{0.40 ± 0.09}\hphantom{*}\hphantom{*} & 0.83 ± 0.50\hphantom{*}\hphantom{*} & 1.24 ± 0.29\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.53 ± 0.01\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 17.45 ± 2.92** & \textbf{79.16 ± 11.67}** & 18.26 ± 4.92** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.10 ± 0.01\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 3.43 ± 1.67** & 186.40 ± 28.06\hphantom{*}\hphantom{*} & \textbf{5.34 ± 2.38}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.51 ± 0.22\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.27 ± 0.33}** & 200.98 ± 28.49\hphantom{*}\hphantom{*} & 6.50 ± 2.01** & \textbf{0.77 ± 0.07}** & 0.07 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM} & Wachter & 57.71 ± 0.47\hphantom{*}\hphantom{*} & 214.08 ± 45.35\hphantom{*}\hphantom{*} & 61.04 ± 2.58\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.07 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 17.43 ± 3.04** & \textbf{83.28 ± 13.26}** & 17.21 ± 4.46** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.16 ± 0.11\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 2.94 ± 1.13** & 194.24 ± 35.41\hphantom{*}\hphantom{*} & \textbf{4.95 ± 1.26}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.51 ± 0.29\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.03 ± 0.20}** & 208.45 ± 34.60\hphantom{*}\hphantom{*} & 6.12 ± 1.91** & \textbf{0.85 ± 0.05}** & 0.09 ± 0.04\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM Ensemble} & Wachter & 56.79 ± 44.68\hphantom{*}\hphantom{*} & 186.19 ± 33.88\hphantom{*}\hphantom{*} & 60.70 ± 44.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.07 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 17.05 ± 2.87** & \textbf{75.93 ± 14.27}** & 17.20 ± 3.15** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.19 ± 0.08\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 2.93 ± 1.24** & 196.75 ± 41.25\hphantom{*}\hphantom{*} & \textbf{4.84 ± 0.60}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.38 ± 0.18\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & Schut & \textbf{1.49 ± 0.87}** & 212.00 ± 41.15\hphantom{*}\hphantom{*} & 6.44 ± 1.34** & \textbf{0.77 ± 0.13}** & 0.12 ± 0.01\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash MLP} & Wachter & 42.97 ± 39.50\hphantom{*}\hphantom{*} & 218.34 ± 53.26\hphantom{*}\hphantom{*} & 45.84 ± 39.39\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.06 ± 0.06}\hphantom{*}\hphantom{*} & 0.50 ± 0.51\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 16.63 ± 2.62** & \textbf{73.86 ± 14.63}** & 17.92 ± 4.17** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.23 ± 0.07\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 3.73 ± 2.36** & 207.21 ± 43.20\hphantom{*}\hphantom{*} & \textbf{5.78 ± 2.10}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.33 ± 0.19\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+ & & Schut & \textbf{1.20 ± 0.47}** & 205.36 ± 32.11\hphantom{*}\hphantom{*} & 7.00 ± 2.15** & \textbf{0.79 ± 0.09}** & 0.12 ± 0.01\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}**\\
+
+\multirow[t]{-16}{*}{\centering\arraybackslash GMSC} & \multirow[t]{-4}{*}{\centering\arraybackslash MLP Ensemble} & Wachter & 69.30 ± 66.00\hphantom{*}\hphantom{*} & 213.71 ± 54.17\hphantom{*}\hphantom{*} & 73.09 ± 64.50\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.06 ± 0.06}\hphantom{*}\hphantom{*} & 0.50 ± 0.51\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 0.75 ± 0.17\hphantom{*}\hphantom{*} & \textbf{0.03 ± 0.06}** & \textbf{0.20 ± 0.08}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.75 ± 0.17\hphantom{*}\hphantom{*} & 0.03 ± 0.06** & 0.20 ± 0.08** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.70 ± 0.16\hphantom{*}\hphantom{*} & 0.16 ± 0.11\hphantom{*}\hphantom{*} & 0.34 ± 0.19\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & \textbf{0.41 ± 0.15}\hphantom{*}\hphantom{*} & 0.19 ± 0.03\hphantom{*}\hphantom{*} & 0.41 ± 0.01** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.36 ± 0.36\hphantom{*}\hphantom{*} & 0.50 ± 0.51\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 1.15 ± 0.35\hphantom{*}\hphantom{*} & 0.39 ± 0.07\hphantom{*}\hphantom{*} & 0.73 ± 0.17\hphantom{*}\hphantom{*} & \textbf{0.25 ± 0.25}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & 0.50 ± 0.13\hphantom{*}\hphantom{*} & 0.18 ± 0.10\hphantom{*}\hphantom{*} & 0.44 ± 0.17\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 0.95 ± 0.16\hphantom{*}\hphantom{*} & \textbf{0.29 ± 0.05}** & 0.23 ± 0.06** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.94 ± 0.16\hphantom{*}\hphantom{*} & 0.29 ± 0.05** & \textbf{0.23 ± 0.07}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.60 ± 0.15\hphantom{*}\hphantom{*} & 0.46 ± 0.05\hphantom{*}\hphantom{*} & 0.28 ± 0.04** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.02 ± 0.10** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & \textbf{0.42 ± 0.14}\hphantom{*}\hphantom{*} & 0.56 ± 0.05\hphantom{*}\hphantom{*} & 0.41 ± 0.01\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.47 ± 0.50\hphantom{*}\hphantom{*} & 0.48 ± 0.50\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 0.77 ± 0.17\hphantom{*}\hphantom{*} & 0.43 ± 0.06*\hphantom{*} & 0.47 ± 0.36\hphantom{*}\hphantom{*} & \textbf{0.20 ± 0.25}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Linearly Separable} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & 0.51 ± 0.15\hphantom{*}\hphantom{*} & 0.51 ± 0.04\hphantom{*}\hphantom{*} & 0.40 ± 0.08\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.59 ± 0.02\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 334.61 ± 46.37\hphantom{*}\hphantom{*} & \textbf{19.28 ± 5.01}** & 314.76 ± 32.36*\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 4.43 ± 0.56\hphantom{*}\hphantom{*} & \textbf{0.98 ± 0.12}\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 170.68 ± 63.26\hphantom{*}\hphantom{*} & 188.70 ± 26.18*\hphantom{*} & \textbf{255.26 ± 41.50}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 4.39 ± 0.91\hphantom{*}\hphantom{*} & 0.96 ± 0.20\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.44 ± 1.60}** & 211.00 ± 27.21\hphantom{*}\hphantom{*} & 286.61 ± 39.85*\hphantom{*} & \textbf{0.99 ± 0.00}** & \textbf{1.08 ± 1.95}*\hphantom{*} & 0.24 ± 0.43\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM} & Wachter & 128.36 ± 14.95\hphantom{*}\hphantom{*} & 222.90 ± 26.56\hphantom{*}\hphantom{*} & 361.88 ± 39.74\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 4.37 ± 0.98\hphantom{*}\hphantom{*} & 0.95 ± 0.21\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 342.64 ± 41.14\hphantom{*}\hphantom{*} & \textbf{15.99 ± 3.06}** & 294.72 ± 30.75** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 2.07 ± 0.06** & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 170.21 ± 58.02\hphantom{*}\hphantom{*} & 173.59 ± 20.65** & \textbf{246.32 ± 37.46}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 2.56 ± 0.83\hphantom{*}\hphantom{*} & 0.93 ± 0.26\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.78 ± 1.02}** & 205.33 ± 24.07\hphantom{*}\hphantom{*} & 287.39 ± 39.33*\hphantom{*} & \textbf{0.99 ± 0.00}** & \textbf{0.32 ± 0.94}** & 0.11 ± 0.31\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM Ensemble} & Wachter & 135.07 ± 16.79\hphantom{*}\hphantom{*} & 217.67 ± 23.78\hphantom{*}\hphantom{*} & 363.23 ± 39.24\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 2.93 ± 0.77\hphantom{*}\hphantom{*} & 0.94 ± 0.23\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 605.17 ± 44.78\hphantom{*}\hphantom{*} & \textbf{41.95 ± 6.50}** & 591.58 ± 36.24\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.57 ± 0.00** & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 146.61 ± 36.96\hphantom{*}\hphantom{*} & 365.82 ± 15.35*\hphantom{*} & \textbf{249.49 ± 41.55}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.62 ± 0.30\hphantom{*}\hphantom{*} & 0.87 ± 0.34\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.95 ± 0.37}** & 382.44 ± 17.81\hphantom{*}\hphantom{*} & 285.98 ± 42.48*\hphantom{*} & \textbf{0.99 ± 0.00}** & \textbf{0.05 ± 0.19}** & 0.06 ± 0.24\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash MLP} & Wachter & 136.08 ± 16.09\hphantom{*}\hphantom{*} & 386.05 ± 16.60\hphantom{*}\hphantom{*} & 361.83 ± 42.18\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.68 ± 0.36\hphantom{*}\hphantom{*} & 0.84 ± 0.36\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 525.87 ± 34.00\hphantom{*}\hphantom{*} & \textbf{31.43 ± 3.91}** & 490.88 ± 27.19\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.29 ± 0.00** & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 146.60 ± 35.64\hphantom{*}\hphantom{*} & 337.74 ± 11.89*\hphantom{*} & \textbf{247.67 ± 38.36}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.39 ± 0.22\hphantom{*}\hphantom{*} & 0.85 ± 0.36\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.98 ± 0.25}** & 359.54 ± 14.52\hphantom{*}\hphantom{*} & 283.99 ± 41.08*\hphantom{*} & \textbf{0.99 ± 0.00}** & \textbf{0.03 ± 0.14}** & 0.06 ± 0.24\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-16}{*}{\centering\arraybackslash MNIST} & \multirow[t]{-4}{*}{\centering\arraybackslash MLP Ensemble} & Wachter & 137.53 ± 18.95\hphantom{*}\hphantom{*} & 360.79 ± 14.39\hphantom{*}\hphantom{*} & 357.73 ± 42.55\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.47 ± 0.64\hphantom{*}\hphantom{*} & 0.80 ± 0.40\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 1.56 ± 0.44\hphantom{*}\hphantom{*} & \textbf{0.31 ± 0.30}*\hphantom{*} & \textbf{1.20 ± 0.15}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}**\\
+
+ & & ECCCo (no CP) & 1.56 ± 0.46\hphantom{*}\hphantom{*} & 0.37 ± 0.30*\hphantom{*} & 1.21 ± 0.17** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}**\\
+
+ & & ECCCo (no EBM) & 0.80 ± 0.25\hphantom{*}\hphantom{*} & 0.91 ± 0.32\hphantom{*}\hphantom{*} & 1.71 ± 0.25\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}**\\
+
+ & & REVISE & 1.04 ± 0.43\hphantom{*}\hphantom{*} & 0.78 ± 0.23\hphantom{*}\hphantom{*} & 1.57 ± 0.26\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & \textbf{1.00 ± 0.00}**\\
+
+ & & Schut & 1.12 ± 0.31\hphantom{*}\hphantom{*} & 0.67 ± 0.27\hphantom{*}\hphantom{*} & 1.50 ± 0.22*\hphantom{*} & \textbf{0.08 ± 0.19}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 0.98 ± 0.14\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & \textbf{0.72 ± 0.24}\hphantom{*}\hphantom{*} & 0.80 ± 0.27\hphantom{*}\hphantom{*} & 1.78 ± 0.24\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.02 ± 0.10\hphantom{*}\hphantom{*} & 0.98 ± 0.14\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 2.18 ± 1.05\hphantom{*}\hphantom{*} & 0.80 ± 0.62\hphantom{*}\hphantom{*} & 1.69 ± 0.40\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.24*\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 2.07 ± 1.15\hphantom{*}\hphantom{*} & \textbf{0.79 ± 0.62}\hphantom{*}\hphantom{*} & 1.68 ± 0.42\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.24*\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 1.25 ± 0.92\hphantom{*}\hphantom{*} & 1.34 ± 0.47\hphantom{*}\hphantom{*} & 1.68 ± 0.47\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.43 ± 0.18\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 0.79 ± 0.19*\hphantom{*} & 1.45 ± 0.44\hphantom{*}\hphantom{*} & \textbf{1.64 ± 0.31}\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.40 ± 0.22\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{0.73 ± 0.25}*\hphantom{*} & 1.45 ± 0.55\hphantom{*}\hphantom{*} & 1.73 ± 0.48\hphantom{*}\hphantom{*} & \textbf{0.31 ± 0.28}*\hphantom{*} & \textbf{0.00 ± 0.00}** & 0.90 ± 0.30\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Moons} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & 1.08 ± 0.83\hphantom{*}\hphantom{*} & 1.32 ± 0.41\hphantom{*}\hphantom{*} & 1.69 ± 0.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.52 ± 0.08\hphantom{*}\hphantom{*} & \textbf{1.00 ± 0.00}\hphantom{*}\hphantom{*}\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table_all_valid.tex b/AAAI/contents/table_all_valid.tex
new file mode 100644
index 0000000000000000000000000000000000000000..3fa029f190735829c3f86cb011d1b1d802f304d6
--- /dev/null
+++ b/AAAI/contents/table_all_valid.tex
@@ -0,0 +1,147 @@
+\begin{table}
+
+\caption{All results for all datasets: sample averages +/- one standard deviation over all valid counterfactuals. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-full-valid} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{ccccccccc}
+\toprule
+Model & Data & Generator & Cost ↓ & Unfaithfulness ↓ & Implausibility ↓ & Redundancy ↑ & Uncertainty ↓ & Validity ↑\\
+\midrule
+ & & ECCCo & 0.74 ± 0.21\hphantom{*}\hphantom{*} & 0.52 ± 0.36\hphantom{*}\hphantom{*} & 1.22 ± 0.46\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.72 ± 0.21\hphantom{*}\hphantom{*} & 0.54 ± 0.39\hphantom{*}\hphantom{*} & 1.21 ± 0.46\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.52 ± 0.15\hphantom{*}\hphantom{*} & 0.70 ± 0.33\hphantom{*}\hphantom{*} & 1.30 ± 0.37\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 1.28 ± 0.14\hphantom{*}\hphantom{*} & \textbf{0.33 ± 0.01}** & \textbf{0.64 ± 0.00}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 1.06 ± 0.43\hphantom{*}\hphantom{*} & 0.54 ± 0.43\hphantom{*}\hphantom{*} & 1.28 ± 0.53\hphantom{*}\hphantom{*} & \textbf{0.26 ± 0.25}*\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & \textbf{0.45 ± 0.15}\hphantom{*}\hphantom{*} & 0.68 ± 0.34\hphantom{*}\hphantom{*} & 1.33 ± 0.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 0.67 ± 0.19\hphantom{*}\hphantom{*} & 0.65 ± 0.53\hphantom{*}\hphantom{*} & 1.17 ± 0.41\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.09 ± 0.19** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.71 ± 0.16\hphantom{*}\hphantom{*} & 0.49 ± 0.35\hphantom{*}\hphantom{*} & 1.19 ± 0.44\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.05 ± 0.16** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.45 ± 0.11\hphantom{*}\hphantom{*} & 0.84 ± 0.51\hphantom{*}\hphantom{*} & 1.23 ± 0.31\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.23*\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 1.24 ± 0.15\hphantom{*}\hphantom{*} & \textbf{0.06 ± 0.01}** & \textbf{0.64 ± 0.00}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 0.57 ± 0.11\hphantom{*}\hphantom{*} & 0.58 ± 0.37\hphantom{*}\hphantom{*} & 1.23 ± 0.43\hphantom{*}\hphantom{*} & \textbf{0.43 ± 0.18}** & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Circles} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & \textbf{0.40 ± 0.09}\hphantom{*}\hphantom{*} & 0.83 ± 0.50\hphantom{*}\hphantom{*} & 1.24 ± 0.29\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.53 ± 0.01\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 17.45 ± 2.92** & \textbf{79.16 ± 11.67}** & 18.26 ± 4.92** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.10 ± 0.01\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 3.43 ± 1.67** & 186.40 ± 28.06\hphantom{*}\hphantom{*} & \textbf{5.34 ± 2.38}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.51 ± 0.22\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.27 ± 0.33}** & 200.98 ± 28.49\hphantom{*}\hphantom{*} & 6.50 ± 2.01** & \textbf{0.77 ± 0.07}** & 0.07 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM} & Wachter & 57.71 ± 0.47\hphantom{*}\hphantom{*} & 214.08 ± 45.35\hphantom{*}\hphantom{*} & 61.04 ± 2.58\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.07 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 17.43 ± 3.04** & \textbf{83.28 ± 13.26}** & 17.21 ± 4.46** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.16 ± 0.11\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 2.94 ± 1.13** & 194.24 ± 35.41\hphantom{*}\hphantom{*} & \textbf{4.95 ± 1.26}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.51 ± 0.29\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.03 ± 0.20}** & 208.45 ± 34.60\hphantom{*}\hphantom{*} & 6.12 ± 1.91** & \textbf{0.85 ± 0.05}** & 0.09 ± 0.04\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM Ensemble} & Wachter & 56.79 ± 44.68\hphantom{*}\hphantom{*} & 186.19 ± 33.88\hphantom{*}\hphantom{*} & 60.70 ± 44.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.07 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 17.05 ± 2.87\hphantom{*}\hphantom{*} & \textbf{75.93 ± 14.27}** & 17.20 ± 3.15\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.19 ± 0.08\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 2.93 ± 1.24*\hphantom{*} & 196.75 ± 41.25\hphantom{*}\hphantom{*} & \textbf{4.84 ± 0.60}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.38 ± 0.18\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.49 ± 0.87}** & 212.00 ± 41.15\hphantom{*}\hphantom{*} & 6.44 ± 1.34\hphantom{*}\hphantom{*} & \textbf{0.77 ± 0.13}** & 0.12 ± 0.01\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash MLP} & Wachter & 4.48 ± 0.18\hphantom{*}\hphantom{*} & 184.03 ± 48.16\hphantom{*}\hphantom{*} & 7.49 ± 0.89\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.12 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 16.63 ± 2.62\hphantom{*}\hphantom{*} & \textbf{73.86 ± 14.63}** & 17.92 ± 4.17\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.23 ± 0.07\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 3.73 ± 2.36\hphantom{*}\hphantom{*} & 207.21 ± 43.20\hphantom{*}\hphantom{*} & \textbf{5.78 ± 2.10}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.33 ± 0.19\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{1.20 ± 0.47}** & 205.36 ± 32.11\hphantom{*}\hphantom{*} & 7.00 ± 2.15*\hphantom{*} & \textbf{0.79 ± 0.09}** & 0.12 ± 0.01\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-16}{*}{\centering\arraybackslash GMSC} & \multirow[t]{-4}{*}{\centering\arraybackslash MLP Ensemble} & Wachter & 4.97 ± 0.47\hphantom{*}\hphantom{*} & 177.20 ± 25.86\hphantom{*}\hphantom{*} & 10.27 ± 3.21\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.11 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 0.75 ± 0.17\hphantom{*}\hphantom{*} & \textbf{0.03 ± 0.06}** & \textbf{0.20 ± 0.08}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.75 ± 0.17\hphantom{*}\hphantom{*} & 0.03 ± 0.06** & 0.20 ± 0.08** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.70 ± 0.16\hphantom{*}\hphantom{*} & 0.16 ± 0.11\hphantom{*}\hphantom{*} & 0.34 ± 0.19\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & \textbf{0.41 ± 0.14}\hphantom{*}\hphantom{*} & 0.15 ± 0.00** & 0.41 ± 0.01** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.72 ± 0.02\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 1.15 ± 0.35\hphantom{*}\hphantom{*} & 0.39 ± 0.07\hphantom{*}\hphantom{*} & 0.73 ± 0.17\hphantom{*}\hphantom{*} & \textbf{0.25 ± 0.25}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & 0.50 ± 0.13\hphantom{*}\hphantom{*} & 0.18 ± 0.10\hphantom{*}\hphantom{*} & 0.44 ± 0.17\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 0.95 ± 0.16\hphantom{*}\hphantom{*} & \textbf{0.29 ± 0.05}** & 0.23 ± 0.06** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 0.94 ± 0.16\hphantom{*}\hphantom{*} & 0.29 ± 0.05** & \textbf{0.23 ± 0.07}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.60 ± 0.15\hphantom{*}\hphantom{*} & 0.46 ± 0.05\hphantom{*}\hphantom{*} & 0.28 ± 0.04** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.02 ± 0.10** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & \textbf{0.39 ± 0.15}\hphantom{*}\hphantom{*} & 0.52 ± 0.04\hphantom{*}\hphantom{*} & 0.41 ± 0.01\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.98 ± 0.00\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 0.77 ± 0.17\hphantom{*}\hphantom{*} & 0.43 ± 0.06*\hphantom{*} & 0.47 ± 0.36\hphantom{*}\hphantom{*} & \textbf{0.20 ± 0.25}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Linearly Separable} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & 0.51 ± 0.15\hphantom{*}\hphantom{*} & 0.51 ± 0.04\hphantom{*}\hphantom{*} & 0.40 ± 0.08\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.59 ± 0.02\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 334.98 ± 46.54\hphantom{*}\hphantom{*} & \textbf{19.27 ± 5.02}** & 314.54 ± 32.54*\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{4.50 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 170.06 ± 62.45\hphantom{*}\hphantom{*} & 188.54 ± 26.22*\hphantom{*} & \textbf{254.32 ± 41.55}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 4.57 ± 0.14\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{7.63 ± 2.55}** & 199.70 ± 28.43\hphantom{*}\hphantom{*} & 273.01 ± 39.60** & \textbf{0.99 ± 0.00}** & 4.56 ± 0.13\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM} & Wachter & 128.13 ± 14.81\hphantom{*}\hphantom{*} & 222.81 ± 26.22\hphantom{*}\hphantom{*} & 361.38 ± 39.55\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 4.58 ± 0.16\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 342.64 ± 41.14\hphantom{*}\hphantom{*} & \textbf{15.99 ± 3.06}** & 294.72 ± 30.75** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{2.07 ± 0.06}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 171.95 ± 58.81\hphantom{*}\hphantom{*} & 173.05 ± 20.38** & \textbf{246.20 ± 37.74}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 2.76 ± 0.45\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{7.96 ± 2.49}** & 186.91 ± 22.98*\hphantom{*} & 264.68 ± 37.58** & \textbf{0.99 ± 0.00}** & 3.02 ± 0.26\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash JEM Ensemble} & Wachter & 134.98 ± 16.95\hphantom{*}\hphantom{*} & 217.37 ± 23.93\hphantom{*}\hphantom{*} & 362.91 ± 39.40\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 3.10 ± 0.31\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 605.17 ± 44.78\hphantom{*}\hphantom{*} & \textbf{41.95 ± 6.50}** & 591.58 ± 36.24\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.57 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 146.76 ± 37.07\hphantom{*}\hphantom{*} & 365.69 ± 14.90*\hphantom{*} & 245.36 ± 39.69** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.72 ± 0.18\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.25 ± 1.31}** & 371.12 ± 19.99\hphantom{*}\hphantom{*} & \textbf{245.11 ± 35.72}** & \textbf{0.99 ± 0.00}** & 0.75 ± 0.23\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-4}{*}{\centering\arraybackslash MLP} & Wachter & 135.08 ± 15.68\hphantom{*}\hphantom{*} & 384.76 ± 16.52\hphantom{*}\hphantom{*} & 359.21 ± 42.03\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.81 ± 0.22\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 525.87 ± 34.00\hphantom{*}\hphantom{*} & \textbf{31.43 ± 3.91}** & 490.88 ± 27.19\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.29 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 146.38 ± 35.18\hphantom{*}\hphantom{*} & 337.21 ± 11.68*\hphantom{*} & \textbf{244.84 ± 37.17}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.45 ± 0.16\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{9.75 ± 1.00}** & 344.60 ± 13.64*\hphantom{*} & 252.53 ± 37.92** & \textbf{0.99 ± 0.00}** & 0.55 ± 0.21\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-16}{*}{\centering\arraybackslash MNIST} & \multirow[t]{-4}{*}{\centering\arraybackslash MLP Ensemble} & Wachter & 134.48 ± 17.69\hphantom{*}\hphantom{*} & 358.51 ± 13.18\hphantom{*}\hphantom{*} & 352.63 ± 39.93\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.58 ± 0.67\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{1-9}
+ & & ECCCo & 1.56 ± 0.44\hphantom{*}\hphantom{*} & \textbf{0.31 ± 0.30}*\hphantom{*} & \textbf{1.20 ± 0.15}** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 1.56 ± 0.46\hphantom{*}\hphantom{*} & 0.37 ± 0.30*\hphantom{*} & 1.21 ± 0.17** & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 0.80 ± 0.25\hphantom{*}\hphantom{*} & 0.91 ± 0.32\hphantom{*}\hphantom{*} & 1.71 ± 0.25\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 1.04 ± 0.43\hphantom{*}\hphantom{*} & 0.78 ± 0.23\hphantom{*}\hphantom{*} & 1.57 ± 0.26\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & 1.13 ± 0.29\hphantom{*}\hphantom{*} & 0.66 ± 0.25\hphantom{*}\hphantom{*} & 1.47 ± 0.10** & \textbf{0.07 ± 0.18}\hphantom{*}\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & \multirow[t]{-6}{*}{\centering\arraybackslash JEM} & Wachter & \textbf{0.73 ± 0.24}\hphantom{*}\hphantom{*} & 0.78 ± 0.23\hphantom{*}\hphantom{*} & 1.75 ± 0.19\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.02 ± 0.11\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\cmidrule{2-9}
+ & & ECCCo & 2.18 ± 1.05\hphantom{*}\hphantom{*} & 0.80 ± 0.62\hphantom{*}\hphantom{*} & 1.69 ± 0.40\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.24*\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no CP) & 2.07 ± 1.15\hphantom{*}\hphantom{*} & \textbf{0.79 ± 0.62}\hphantom{*}\hphantom{*} & 1.68 ± 0.42\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.15 ± 0.24*\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & ECCCo (no EBM) & 1.25 ± 0.92\hphantom{*}\hphantom{*} & 1.34 ± 0.47\hphantom{*}\hphantom{*} & 1.68 ± 0.47\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.43 ± 0.18\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & REVISE & 0.79 ± 0.19*\hphantom{*} & 1.45 ± 0.44\hphantom{*}\hphantom{*} & 1.64 ± 0.31\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.40 ± 0.22\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+ & & Schut & \textbf{0.78 ± 0.17}*\hphantom{*} & 1.39 ± 0.50\hphantom{*}\hphantom{*} & \textbf{1.59 ± 0.26}\hphantom{*}\hphantom{*} & \textbf{0.28 ± 0.25}*\hphantom{*} & \textbf{0.00 ± 0.00}** & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+
+\multirow[t]{-12}{*}{\centering\arraybackslash Moons} & \multirow[t]{-6}{*}{\centering\arraybackslash MLP} & Wachter & 1.08 ± 0.83\hphantom{*}\hphantom{*} & 1.32 ± 0.41\hphantom{*}\hphantom{*} & 1.69 ± 0.32\hphantom{*}\hphantom{*} & 0.00 ± 0.00\hphantom{*}\hphantom{*} & 0.52 ± 0.08\hphantom{*}\hphantom{*} & 1.00 ± 0.00\hphantom{*}\hphantom{*}\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table_ebm_params.tex b/AAAI/contents/table_ebm_params.tex
new file mode 100644
index 0000000000000000000000000000000000000000..e7fe1e0907384a1c0cf18ccdcef83720d6b91d45
--- /dev/null
+++ b/AAAI/contents/table_ebm_params.tex
@@ -0,0 +1,17 @@
+\begin{table}
+
+\caption{EBM hyperparemeter choices for our experiments. \label{tab:ebmparams} \newline}
+\centering
+\fontsize{8}{10}\selectfont
+\begin{tabular}[t]{rrrr}
+\toprule
+Dataset & SGLD Steps & Batch Size & $\lambda$\\
+\midrule
+Linearly Separable & 30 & 50 & 0.10\\
+Moons & 30 & 10 & 0.10\\
+Circles & 20 & 100 & 0.01\\
+MNIST & 25 & 10 & 0.01\\
+GMSC & 30 & 10 & 0.10\\
+\bottomrule
+\end{tabular}
+\end{table}
diff --git a/AAAI/contents/table_gen_params.tex b/AAAI/contents/table_gen_params.tex
new file mode 100644
index 0000000000000000000000000000000000000000..84b89401bdaebddd0a0c92778fe91d5cc0122d2b
--- /dev/null
+++ b/AAAI/contents/table_gen_params.tex
@@ -0,0 +1,17 @@
+\begin{table}
+
+\caption{Generator hyperparameters. \label{tab:genparams} \newline}
+\centering
+\fontsize{8}{10}\selectfont
+\begin{tabular}[t]{rrrrr}
+\toprule
+Dataset & $\eta$ & $\lambda_1$ & $\lambda_2$ & $\lambda_3$\\
+\midrule
+Linearly Separable & 0.01 & 0.25 & 0.75 & 0.75\\
+Moons & 0.05 & 0.25 & 0.75 & 0.75\\
+Circles & 0.01 & 0.25 & 0.75 & 0.75\\
+MNIST & 0.10 & 0.10 & 0.25 & 0.25\\
+GMSC & 0.05 & 0.10 & 0.50 & 0.50\\
+\bottomrule
+\end{tabular}
+\end{table}
diff --git a/AAAI/contents/table_params.tex b/AAAI/contents/table_params.tex
new file mode 100644
index 0000000000000000000000000000000000000000..d0ccc5a3ef1f60a4ff3f5d39f629e41ae4d3d46e
--- /dev/null
+++ b/AAAI/contents/table_params.tex
@@ -0,0 +1,19 @@
+\begin{table}
+
+\caption{Paremeter choices for our experiments. \label{tab:params} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{rrrrrrrr}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{4}{c}{Network Architecture} & \multicolumn{2}{c}{Training} \\
+\cmidrule(l{3pt}r{3pt}){3-6} \cmidrule(l{3pt}r{3pt}){7-8}
+Dataset & Sample Size & Hidden Units & Hidden Layers & Activation & Ensemble Size & Epochs & Batch Size\\
+\midrule
+Linearly Separable & 1000 & 16 & 3 & swish & 5 & 100 & 100\\
+Moons & 2500 & 32 & 3 & relu & 5 & 500 & 128\\
+Circles & 1000 & 32 & 3 & swish & 5 & 100 & 100\\
+MNIST & 10000 & 128 & 1 & swish & 5 & 100 & 128\\
+GMSC & 13370 & 128 & 2 & swish & 5 & 100 & 250\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table_perf.tex b/AAAI/contents/table_perf.tex
new file mode 100644
index 0000000000000000000000000000000000000000..737a98e599580f7d0a38c2c1ec377b9fbed577a8
--- /dev/null
+++ b/AAAI/contents/table_perf.tex
@@ -0,0 +1,41 @@
+\begin{table}
+
+\caption{Various standard performance metrics for our different models grouped by dataset. \label{tab:perf} \newline}
+\centering
+\fontsize{8}{10}\selectfont
+\begin{tabular}[t]{rrrrr}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{3}{c}{Performance Metrics} \\
+\cmidrule(l{3pt}r{3pt}){3-5}
+Dataset & Model & Accuracy & Precision & F1-Score\\
+\midrule
+ & JEM & 0.99 & 0.99 & 0.99\\
+
+\multirow[t]{-2}{*}{\raggedleft\arraybackslash Linearly Separable} & MLP & 0.99 & 0.99 & 0.99\\
+\cmidrule{1-5}
+ & JEM & 1.00 & 1.00 & 1.00\\
+
+\multirow[t]{-2}{*}{\raggedleft\arraybackslash Moons} & MLP & 1.00 & 1.00 & 1.00\\
+\cmidrule{1-5}
+ & JEM & 0.98 & 0.98 & 0.98\\
+
+\multirow[t]{-2}{*}{\raggedleft\arraybackslash Circles} & MLP & 1.00 & 1.00 & 1.00\\
+\cmidrule{1-5}
+ & JEM & 0.83 & 0.84 & 0.83\\
+
+ & JEM Ensemble & 0.90 & 0.90 & 0.89\\
+
+ & MLP & 0.95 & 0.95 & 0.95\\
+
+\multirow[t]{-4}{*}{\raggedleft\arraybackslash MNIST} & MLP Ensemble & 0.95 & 0.95 & 0.95\\
+\cmidrule{1-5}
+ & JEM & 0.73 & 0.75 & 0.73\\
+
+ & JEM Ensemble & 0.73 & 0.75 & 0.73\\
+
+ & MLP & 0.75 & 0.75 & 0.75\\
+
+\multirow[t]{-4}{*}{\raggedleft\arraybackslash GMSC} & MLP Ensemble & 0.75 & 0.75 & 0.75\\
+\bottomrule
+\end{tabular}
+\end{table}
diff --git a/AAAI/contents/table_real.tex b/AAAI/contents/table_real.tex
new file mode 100644
index 0000000000000000000000000000000000000000..43365ededec415bb645a1224b72ed5a590a90529
--- /dev/null
+++ b/AAAI/contents/table_real.tex
@@ -0,0 +1,45 @@
+\begin{table}
+
+\caption{Results for real-world datasets. Standard deviations across samples are shown in parentheses. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-real} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{>{}cccccccc}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{2}{c}{California Housing} & \multicolumn{2}{c}{GMSC} & \multicolumn{2}{c}{MNIST} \\
+\cmidrule(l{3pt}r{3pt}){3-4} \cmidrule(l{3pt}r{3pt}){5-6} \cmidrule(l{3pt}r{3pt}){7-8}
+Model & Generator & Non-conformity ↓ & Implausibility ↓ & Non-conformity ↓ & Implausibility ↓ & Non-conformity ↓ & Implausibility ↓\\
+\midrule
+ & ECCCo & \textbf{236.79 (51.16)} & 39.78 (3.18) & \textbf{41.65 (17.24)**} & 40.57 (8.74)** & \textbf{116.09 (30.70)**} & 281.33 (41.51)**\\
+
+ & REVISE & 284.51 (52.74) & \textbf{5.58 (0.81)**} & 74.89 (15.82)** & \textbf{6.01 (5.75)**} & 348.74 (65.65)** & \textbf{246.69 (36.69)**}\\
+
+ & Schut & 263.55 (60.56) & 8.00 (2.03) & 76.23 (15.54)** & 6.02 (0.72)** & 355.58 (64.84)** & 270.06 (40.41)**\\
+
+\multirow{-4}{*}{\centering\arraybackslash \textbf{JEM}} & Wachter & 274.55 (51.17) & 7.32 (1.80) & 146.02 (64.48) & 128.93 (74.00) & 694.08 (50.86) & 630.99 (33.01)\\
+\cmidrule{1-8}
+ & ECCCo & \textbf{249.44 (58.53)} & 35.09 (5.56) & \textbf{26.55 (12.94)**} & 33.65 (8.33)** & \textbf{89.89 (27.26)**} & 240.59 (37.41)**\\
+
+ & REVISE & 268.45 (66.87) & \textbf{5.44 (0.74)**} & 52.47 (14.12)** & 6.69 (3.37)** & 292.52 (53.13)** & \textbf{240.50 (35.73)**}\\
+
+ & Schut & 279.38 (63.23) & 7.64 (1.47) & 56.34 (15.00)** & \textbf{6.27 (1.06)**} & 319.45 (59.02)** & 266.80 (40.46)**\\
+
+\multirow{-4}{*}{\centering\arraybackslash \textbf{JEM Ensemble}} & Wachter & 268.59 (68.66) & 7.16 (1.46) & 125.72 (70.80) & 126.55 (93.75) & 582.52 (58.46) & 543.90 (44.24)\\
+\cmidrule{1-8}
+ & ECCCo & \textbf{230.92 (48.86)} & 37.53 (5.40) & \textbf{46.90 (15.80)**} & 37.78 (8.40)** & \textbf{212.45 (36.70)**} & 649.63 (58.80)\\
+
+ & REVISE & 281.10 (53.01) & \textbf{5.34 (0.67)**} & 81.08 (19.53)** & \textbf{4.60 (0.72)**} & 839.79 (77.14)* & \textbf{244.33 (38.69)**}\\
+
+ & Schut & 285.12 (56.00) & 6.48 (1.18)** & 90.67 (20.80)** & 5.56 (0.81)** & 842.80 (82.01)* & 264.94 (42.18)**\\
+
+\multirow{-4}{*}{\centering\arraybackslash \textbf{MLP}} & Wachter & 262.50 (56.87) & 9.21 (10.41) & 191.68 (30.86) & 200.23 (15.05) & 982.32 (61.81) & 561.23 (45.08)\\
+\cmidrule{1-8}
+ & ECCCo & \textbf{212.47 (59.27)*} & 38.17 (6.18) & \textbf{74.65 (144.69)*} & 71.87 (145.19) & \textbf{162.21 (36.21)**} & 587.65 (95.01)\\
+
+ & REVISE & 284.65 (49.52) & \textbf{5.64 (1.13)*} & 80.90 (14.59)** & \textbf{5.20 (1.52)**} & 741.30 (125.98)* & \textbf{242.76 (41.16)**}\\
+
+ & Schut & 269.19 (46.08) & 7.30 (1.94) & 85.63 (19.15)** & 6.00 (0.99)** & 754.35 (132.26) & 266.94 (42.55)**\\
+
+\multirow{-4}{*}{\centering\arraybackslash \textbf{MLP Ensemble}} & Wachter & 278.09 (73.65) & 7.32 (1.75) & 220.05 (17.41) & 203.65 (14.77) & 871.09 (92.36) & 536.24 (48.73)\\
+\bottomrule
+\end{tabular}}
+\end{table}
diff --git a/AAAI/contents/table_synth.tex b/AAAI/contents/table_synth.tex
new file mode 100644
index 0000000000000000000000000000000000000000..9e966e12acb047725654086204112483da501ff6
--- /dev/null
+++ b/AAAI/contents/table_synth.tex
@@ -0,0 +1,37 @@
+\begin{table}
+
+\caption{Results for synthetic datasets. Standard deviations across samples are shown in parentheses. Best outcomes are highlighted in bold. Asterisks indicate that the given value is more than one (*) or two (**) standard deviations away from the baseline (Wachter). \label{tab:results-synth} \newline}
+\centering
+\resizebox{\linewidth}{!}{
+\begin{tabular}[t]{>{}cccccccc}
+\toprule
+\multicolumn{2}{c}{ } & \multicolumn{2}{c}{Circles} & \multicolumn{2}{c}{Linearly Separable} & \multicolumn{2}{c}{Moons} \\
+\cmidrule(l{3pt}r{3pt}){3-4} \cmidrule(l{3pt}r{3pt}){5-6} \cmidrule(l{3pt}r{3pt}){7-8}
+Model & Generator & Non-conformity ↓ & Implausibility ↓ & Non-conformity ↓ & Implausibility ↓ & Non-conformity ↓ & Implausibility ↓\\
+\midrule
+ & ECCCo & \textbf{0.63 (1.58)} & 1.44 (1.37) & 0.10 (0.06)** & 0.19 (0.03)** & \textbf{0.57 (0.58)**} & \textbf{1.29 (0.21)*}\\
+
+ & ECCCo (no CP) & 0.64 (1.61) & 1.45 (1.38) & \textbf{0.10 (0.07)**} & \textbf{0.19 (0.03)**} & 0.63 (0.64)* & 1.30 (0.21)*\\
+
+ & ECCCo (no EBM) & 1.41 (1.51) & 1.50 (1.38) & 0.37 (0.28) & 0.38 (0.26) & 1.73 (1.34) & 1.73 (1.42)\\
+
+ & REVISE & 0.96 (0.32)* & \textbf{0.95 (0.32)*} & 0.41 (0.02)** & 0.41 (0.01)** & 1.59 (0.55) & 1.55 (0.20)\\
+
+ & Schut & 0.99 (0.80) & 1.28 (0.53) & 0.66 (0.23) & 0.66 (0.22) & 1.55 (0.61) & 1.42 (0.16)*\\
+
+\multirow{-6}{*}{\centering\arraybackslash \textbf{JEM}} & Wachter & 1.41 (1.50) & 1.51 (1.35) & 0.44 (0.16) & 0.44 (0.15) & 1.77 (0.48) & 1.67 (0.15)\\
+\cmidrule{1-8}
+ & ECCCo & \textbf{0.37 (0.65)**} & 1.30 (0.68) & \textbf{0.03 (0.02)**} & 0.69 (0.10) & 1.68 (1.74) & 2.02 (0.86)\\
+
+ & ECCCo (no CP) & 0.50 (0.85)* & 1.28 (0.66) & \textbf{0.03 (0.02)**} & 0.68 (0.10) & \textbf{1.34 (1.66)} & 2.11 (0.88)\\
+
+ & ECCCo (no EBM) & 2.00 (1.46) & 1.83 (1.00) & 1.25 (0.87) & 1.84 (1.10) & 2.98 (1.89) & 2.29 (1.75)\\
+
+ & REVISE & 1.16 (1.05) & \textbf{0.95 (0.32)*} & 1.10 (0.10) & \textbf{0.40 (0.01)**} & 2.46 (1.05) & \textbf{1.54 (0.27)*}\\
+
+ & Schut & 1.60 (1.15) & 1.24 (0.44) & 0.81 (0.10)* & 0.47 (0.24) & 2.71 (1.15) & 1.62 (0.42)\\
+
+\multirow{-6}{*}{\centering\arraybackslash \textbf{MLP}} & Wachter & 1.67 (1.05) & 1.31 (0.43) & 0.94 (0.11) & 0.44 (0.15) & 2.95 (1.42) & 1.84 (1.33)\\
+\bottomrule
+\end{tabular}}
+\end{table}