Conference Paper/Proceeding/Abstract 565 views 82 downloads
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers
Lecture Notes in Computer Science, Volume: 13515, Pages: 411 - 420
Swansea University Authors:
Xiuyi Fan, Monika Seisenberger
-
PDF | Accepted Manuscript
Download (303.31KB)
DOI (Published version): 10.1007/978-3-031-16564-1_39
Abstract
Local explanations aim to provide transparency for individual instances and their associated predictions. The need for local explanations is prominent for high-risk domains such as finance, law and health care. We propose a new model-agnostic framework for local explanations ``Polynomial Adaptive Lo...
Published in: | Lecture Notes in Computer Science |
---|---|
ISBN: | 9783031165634 9783031165641 |
ISSN: | 0302-9743 1611-3349 |
Published: |
Cham
Springer International Publishing
2022
|
Online Access: |
Check full text
|
URI: | https://cronfa.swan.ac.uk/Record/cronfa60605 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
first_indexed |
2022-07-22T18:41:59Z |
---|---|
last_indexed |
2023-01-13T19:20:51Z |
id |
cronfa60605 |
recordtype |
SURis |
fullrecord |
<?xml version="1.0"?><rfc1807 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"><datestamp>2022-11-17T13:59:15.6954271</datestamp><bib-version>v2</bib-version><id>60605</id><entry>2022-07-22</entry><title>Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers</title><swanseaauthors><author><sid>a88a07c43b3e80f27cb96897d1bc2534</sid><firstname>Xiuyi</firstname><surname>Fan</surname><name>Xiuyi Fan</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>d035399b2b324a63fe472ce0344653e0</sid><ORCID>0000-0002-2226-386X</ORCID><firstname>Monika</firstname><surname>Seisenberger</surname><name>Monika Seisenberger</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2022-07-22</date><abstract>Local explanations aim to provide transparency for individual instances and their associated predictions. The need for local explanations is prominent for high-risk domains such as finance, law and health care. We propose a new model-agnostic framework for local explanations ``Polynomial Adaptive Local Explanations (PALE)", to combat the lack of transparency of predictions through adaptive local models. We aim to explore explanations of predictions by assessing the impact of instantaneous rate of change in each feature and the association with the resulting prediction of the local model. PALE optimises a complex black-box model and the local explanation models for each instance, providing two forms of explanations, one provided by a localised derivative of an adapting polynomial, thus emphasising instance specificity, and the latter a core interpretable logistic regression model.</abstract><type>Conference Paper/Proceeding/Abstract</type><journal>Lecture Notes in Computer Science</journal><volume>13515</volume><journalNumber/><paginationStart>411</paginationStart><paginationEnd>420</paginationEnd><publisher>Springer International Publishing</publisher><placeOfPublication>Cham</placeOfPublication><isbnPrint>9783031165634</isbnPrint><isbnElectronic>9783031165641</isbnElectronic><issnPrint>0302-9743</issnPrint><issnElectronic>1611-3349</issnElectronic><keywords>Explainable AI, Feature Importance, Electronic health records.</keywords><publishedDay>26</publishedDay><publishedMonth>9</publishedMonth><publishedYear>2022</publishedYear><publishedDate>2022-09-26</publishedDate><doi>10.1007/978-3-031-16564-1_39</doi><url/><notes/><college>COLLEGE NANME</college><CollegeCode>COLLEGE CODE</CollegeCode><institution>Swansea University</institution><apcterm>Other</apcterm><funders/><projectreference/><lastEdited>2022-11-17T13:59:15.6954271</lastEdited><Created>2022-07-22T19:15:28.1789458</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>Jamie</firstname><surname>Duell</surname><orcid>0000-0002-8837-7843</orcid><order>1</order></author><author><firstname>Xiuyi</firstname><surname>Fan</surname><order>2</order></author><author><firstname>Monika</firstname><surname>Seisenberger</surname><orcid>0000-0002-2226-386X</orcid><order>3</order></author></authors><documents><document><filename>60605__24716__49ec15ecd8334b769366615f40f80c4f.pdf</filename><originalFilename>ISMIS_2022_Accepted_PALE.pdf</originalFilename><uploaded>2022-07-22T19:37:57.3180842</uploaded><type>Output</type><contentLength>310589</contentLength><contentType>application/pdf</contentType><version>Accepted Manuscript</version><cronfaStatus>true</cronfaStatus><embargoDate>2022-09-26T00:00:00.0000000</embargoDate><copyrightCorrect>true</copyrightCorrect><language>eng</language></document></documents><OutputDurs><OutputDur><Id>136</Id><IsDataAvailableOnline xsi:nil="true"/><DataNotAvailableOnlineReasonId xsi:nil="true"/><IsDurRestrictions xsi:nil="true"/><DurRestrictionReasonId xsi:nil="true"/><DurEmbargoDate xsi:nil="true"/></OutputDur></OutputDurs></rfc1807> |
spelling |
2022-11-17T13:59:15.6954271 v2 60605 2022-07-22 Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers a88a07c43b3e80f27cb96897d1bc2534 Xiuyi Fan Xiuyi Fan true false d035399b2b324a63fe472ce0344653e0 0000-0002-2226-386X Monika Seisenberger Monika Seisenberger true false 2022-07-22 Local explanations aim to provide transparency for individual instances and their associated predictions. The need for local explanations is prominent for high-risk domains such as finance, law and health care. We propose a new model-agnostic framework for local explanations ``Polynomial Adaptive Local Explanations (PALE)", to combat the lack of transparency of predictions through adaptive local models. We aim to explore explanations of predictions by assessing the impact of instantaneous rate of change in each feature and the association with the resulting prediction of the local model. PALE optimises a complex black-box model and the local explanation models for each instance, providing two forms of explanations, one provided by a localised derivative of an adapting polynomial, thus emphasising instance specificity, and the latter a core interpretable logistic regression model. Conference Paper/Proceeding/Abstract Lecture Notes in Computer Science 13515 411 420 Springer International Publishing Cham 9783031165634 9783031165641 0302-9743 1611-3349 Explainable AI, Feature Importance, Electronic health records. 26 9 2022 2022-09-26 10.1007/978-3-031-16564-1_39 COLLEGE NANME COLLEGE CODE Swansea University Other 2022-11-17T13:59:15.6954271 2022-07-22T19:15:28.1789458 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Jamie Duell 0000-0002-8837-7843 1 Xiuyi Fan 2 Monika Seisenberger 0000-0002-2226-386X 3 60605__24716__49ec15ecd8334b769366615f40f80c4f.pdf ISMIS_2022_Accepted_PALE.pdf 2022-07-22T19:37:57.3180842 Output 310589 application/pdf Accepted Manuscript true 2022-09-26T00:00:00.0000000 true eng 136 |
title |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
spellingShingle |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers Xiuyi Fan Monika Seisenberger |
title_short |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
title_full |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
title_fullStr |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
title_full_unstemmed |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
title_sort |
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers |
author_id_str_mv |
a88a07c43b3e80f27cb96897d1bc2534 d035399b2b324a63fe472ce0344653e0 |
author_id_fullname_str_mv |
a88a07c43b3e80f27cb96897d1bc2534_***_Xiuyi Fan d035399b2b324a63fe472ce0344653e0_***_Monika Seisenberger |
author |
Xiuyi Fan Monika Seisenberger |
author2 |
Jamie Duell Xiuyi Fan Monika Seisenberger |
format |
Conference Paper/Proceeding/Abstract |
container_title |
Lecture Notes in Computer Science |
container_volume |
13515 |
container_start_page |
411 |
publishDate |
2022 |
institution |
Swansea University |
isbn |
9783031165634 9783031165641 |
issn |
0302-9743 1611-3349 |
doi_str_mv |
10.1007/978-3-031-16564-1_39 |
publisher |
Springer International Publishing |
college_str |
Faculty of Science and Engineering |
hierarchytype |
|
hierarchy_top_id |
facultyofscienceandengineering |
hierarchy_top_title |
Faculty of Science and Engineering |
hierarchy_parent_id |
facultyofscienceandengineering |
hierarchy_parent_title |
Faculty of Science and Engineering |
department_str |
School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science |
document_store_str |
1 |
active_str |
0 |
description |
Local explanations aim to provide transparency for individual instances and their associated predictions. The need for local explanations is prominent for high-risk domains such as finance, law and health care. We propose a new model-agnostic framework for local explanations ``Polynomial Adaptive Local Explanations (PALE)", to combat the lack of transparency of predictions through adaptive local models. We aim to explore explanations of predictions by assessing the impact of instantaneous rate of change in each feature and the association with the resulting prediction of the local model. PALE optimises a complex black-box model and the local explanation models for each instance, providing two forms of explanations, one provided by a localised derivative of an adapting polynomial, thus emphasising instance specificity, and the latter a core interpretable logistic regression model. |
published_date |
2022-09-26T04:18:51Z |
_version_ |
1763754252738494464 |
score |
11.021648 |