Towards formal fairness in machine learning

Alexey Ignatiev, Martin C. Cooper, Mohamed Siala, Emmanuel Hebrard, Joao Marques-Silva

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

Abstract

One of the challenges of deploying machine learning (ML) systems is fairness. Datasets often include sensitive features, which ML algorithms may unwittingly use to create models that exhibit unfairness. Past work on fairness offers no formal guarantees in their results. This paper proposes to exploit formal reasoning methods to tackle fairness. Starting from an intuitive criterion for fairness of an ML model, the paper formalises it, and shows how fairness can be represented as a decision problem, given some logic representation of an ML model. The same criterion can also be applied to assessing bias in training data. Moreover, we propose a reasonable set of axiomatic properties which no other definition of dataset bias can satisfy. The paper also investigates the relationship between fairness and explainability, and shows that approaches for computing explanations can serve to assess fairness of particular predictions. Finally, the paper proposes SAT-based approaches for learning fair ML models, even when the training data exhibits bias, and reports experimental trials.

Original languageEnglish
Title of host publicationPrinciples and Practice of Constraint Programming
Subtitle of host publication26th International Conference, CP 2020 Louvain-la-Neuve, Belgium, September 7–11, 2020 Proceedings
EditorsHelmut Simonis
Place of PublicationCham Switzerland
PublisherSpringer
Pages846-867
Number of pages22
ISBN (Electronic)9783030584757
ISBN (Print)9783030584740
DOIs
Publication statusPublished - 2020
EventInternational Conference on Principles and Practice of Constraint Programming 2020 - Louvain-la-Neuve, Belgium
Duration: 7 Sep 202011 Sep 2020
Conference number: 26th
https://link.springer.com/book/10.1007/978-3-030-58475-7 (Proceedings)
https://cp2020.a4cp.org (Website)

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume12333
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Conference on Principles and Practice of Constraint Programming 2020
Abbreviated titleCP2020
CountryBelgium
CityLouvain-la-Neuve
Period7/09/2011/09/20
Internet address

Cite this

Ignatiev, A., Cooper, M. C., Siala, M., Hebrard, E., & Marques-Silva, J. (2020). Towards formal fairness in machine learning. In H. Simonis (Ed.), Principles and Practice of Constraint Programming : 26th International Conference, CP 2020 Louvain-la-Neuve, Belgium, September 7–11, 2020 Proceedings (pp. 846-867). (Lecture Notes in Computer Science ; Vol. 12333). Springer. https://doi.org/10.1007/978-3-030-58475-7_49