ICIE 1.0: a novel tool for interactive contextual interaction explanations

Simon B. van der Zon, Wouter Duivesteijn, Werner van Ipenburg, Jan Veldsink, Mykola Pechenizkiy

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

2 Downloads (Pure)

Uittreksel

With the rise of new laws around privacy and awareness, explanation of automated decision making becomes increasingly important. Nowadays, machine learning models are used to aid experts in domains such as banking and insurance to find suspicious transactions, approve loans and credit card applications. Companies using such systems have to be able to provide the rationale behind their decisions; blindly relying on the trained model is not sufficient. There are currently a number of methods that provide insights in models and their decisions, but often they are either good at showing global or local behavior. Global behavior is often too complex to visualize or comprehend, so approximations are shown, and visualizing local behavior is often misleading as it is difficult to define what local exactly means (i.e. our methods don’t “know” how easily a feature-value can be changed; which ones are flexible, and which ones are static). We introduce the ICIE framework (Interactive Contextual Interaction Explanations) which enables users to view explanations of individual instances under different contexts. We will see that various contexts for the same case lead to different explanations, revealing different feature interactions.

Originele taal-2Engels
TitelECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings
RedacteurenAnna Monreale, Carlos Alzate
Plaats van productieCham
UitgeverijSpringer
Pagina's81-94
Aantal pagina's14
ISBN van elektronische versie978-3-030-13463-1
ISBN van geprinte versie978-3-030-13462-4
DOI's
StatusGepubliceerd - 1 jan 2019
Evenement3rd Workshop on Mining Data for Financial Applications, MIDAS 2018 and 2nd International Workshop on Personal Analytics and Privacy, PAP 2018 held at 18th European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2018 - Dublin, Ierland
Duur: 10 sep 201814 sep 2018

Publicatie series

NaamLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11054 LNAI
ISSN van geprinte versie0302-9743
ISSN van elektronische versie1611-3349

Congres

Congres3rd Workshop on Mining Data for Financial Applications, MIDAS 2018 and 2nd International Workshop on Personal Analytics and Privacy, PAP 2018 held at 18th European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2018
LandIerland
StadDublin
Periode10/09/1814/09/18

Vingerafdruk

Interaction
Banking
Insurance
Privacy
Transactions
Learning systems
Machine Learning
Decision making
Decision Making
Model
Sufficient
Approximation
Industry
Context
Awareness
Framework

Citeer dit

van der Zon, S. B., Duivesteijn, W., van Ipenburg, W., Veldsink, J., & Pechenizkiy, M. (2019). ICIE 1.0: a novel tool for interactive contextual interaction explanations. In A. Monreale, & C. Alzate (editors), ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings (blz. 81-94). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11054 LNAI). Cham: Springer. https://doi.org/10.1007/978-3-030-13463-1_6
van der Zon, Simon B. ; Duivesteijn, Wouter ; van Ipenburg, Werner ; Veldsink, Jan ; Pechenizkiy, Mykola. / ICIE 1.0 : a novel tool for interactive contextual interaction explanations. ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings. redacteur / Anna Monreale ; Carlos Alzate. Cham : Springer, 2019. blz. 81-94 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{b86f0dbc197a42119c140adca414c04f,
title = "ICIE 1.0: a novel tool for interactive contextual interaction explanations",
abstract = "With the rise of new laws around privacy and awareness, explanation of automated decision making becomes increasingly important. Nowadays, machine learning models are used to aid experts in domains such as banking and insurance to find suspicious transactions, approve loans and credit card applications. Companies using such systems have to be able to provide the rationale behind their decisions; blindly relying on the trained model is not sufficient. There are currently a number of methods that provide insights in models and their decisions, but often they are either good at showing global or local behavior. Global behavior is often too complex to visualize or comprehend, so approximations are shown, and visualizing local behavior is often misleading as it is difficult to define what local exactly means (i.e. our methods don’t “know” how easily a feature-value can be changed; which ones are flexible, and which ones are static). We introduce the ICIE framework (Interactive Contextual Interaction Explanations) which enables users to view explanations of individual instances under different contexts. We will see that various contexts for the same case lead to different explanations, revealing different feature interactions.",
keywords = "Awareness, Explanations, Feature contributions, Feature interactions, Model transparency, Responsible analytics, Trust",
author = "{van der Zon}, {Simon B.} and Wouter Duivesteijn and {van Ipenburg}, Werner and Jan Veldsink and Mykola Pechenizkiy",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-13463-1_6",
language = "English",
isbn = "978-3-030-13462-4",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "81--94",
editor = "Anna Monreale and Carlos Alzate",
booktitle = "ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings",
address = "Germany",

}

van der Zon, SB, Duivesteijn, W, van Ipenburg, W, Veldsink, J & Pechenizkiy, M 2019, ICIE 1.0: a novel tool for interactive contextual interaction explanations. in A Monreale & C Alzate (redactie), ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11054 LNAI, Springer, Cham, blz. 81-94, Dublin, Ierland, 10/09/18. https://doi.org/10.1007/978-3-030-13463-1_6

ICIE 1.0 : a novel tool for interactive contextual interaction explanations. / van der Zon, Simon B.; Duivesteijn, Wouter; van Ipenburg, Werner; Veldsink, Jan; Pechenizkiy, Mykola.

ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings. redactie / Anna Monreale; Carlos Alzate. Cham : Springer, 2019. blz. 81-94 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11054 LNAI).

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

TY - GEN

T1 - ICIE 1.0

T2 - a novel tool for interactive contextual interaction explanations

AU - van der Zon, Simon B.

AU - Duivesteijn, Wouter

AU - van Ipenburg, Werner

AU - Veldsink, Jan

AU - Pechenizkiy, Mykola

PY - 2019/1/1

Y1 - 2019/1/1

N2 - With the rise of new laws around privacy and awareness, explanation of automated decision making becomes increasingly important. Nowadays, machine learning models are used to aid experts in domains such as banking and insurance to find suspicious transactions, approve loans and credit card applications. Companies using such systems have to be able to provide the rationale behind their decisions; blindly relying on the trained model is not sufficient. There are currently a number of methods that provide insights in models and their decisions, but often they are either good at showing global or local behavior. Global behavior is often too complex to visualize or comprehend, so approximations are shown, and visualizing local behavior is often misleading as it is difficult to define what local exactly means (i.e. our methods don’t “know” how easily a feature-value can be changed; which ones are flexible, and which ones are static). We introduce the ICIE framework (Interactive Contextual Interaction Explanations) which enables users to view explanations of individual instances under different contexts. We will see that various contexts for the same case lead to different explanations, revealing different feature interactions.

AB - With the rise of new laws around privacy and awareness, explanation of automated decision making becomes increasingly important. Nowadays, machine learning models are used to aid experts in domains such as banking and insurance to find suspicious transactions, approve loans and credit card applications. Companies using such systems have to be able to provide the rationale behind their decisions; blindly relying on the trained model is not sufficient. There are currently a number of methods that provide insights in models and their decisions, but often they are either good at showing global or local behavior. Global behavior is often too complex to visualize or comprehend, so approximations are shown, and visualizing local behavior is often misleading as it is difficult to define what local exactly means (i.e. our methods don’t “know” how easily a feature-value can be changed; which ones are flexible, and which ones are static). We introduce the ICIE framework (Interactive Contextual Interaction Explanations) which enables users to view explanations of individual instances under different contexts. We will see that various contexts for the same case lead to different explanations, revealing different feature interactions.

KW - Awareness

KW - Explanations

KW - Feature contributions

KW - Feature interactions

KW - Model transparency

KW - Responsible analytics

KW - Trust

UR - http://www.scopus.com/inward/record.url?scp=85063448784&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-13463-1_6

DO - 10.1007/978-3-030-13463-1_6

M3 - Conference contribution

AN - SCOPUS:85063448784

SN - 978-3-030-13462-4

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 81

EP - 94

BT - ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings

A2 - Monreale, Anna

A2 - Alzate, Carlos

PB - Springer

CY - Cham

ER -

van der Zon SB, Duivesteijn W, van Ipenburg W, Veldsink J, Pechenizkiy M. ICIE 1.0: a novel tool for interactive contextual interaction explanations. In Monreale A, Alzate C, redacteurs, ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018, Proceedings. Cham: Springer. 2019. blz. 81-94. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-13463-1_6