Timezone: »
Equalized odds is a statistical notion of fairness in machine learning that ensures that classification algorithms do not discriminate against protected groups. We extend equalized odds to the setting of cardinality-constrained fair classification, where we have a bounded amount of a resource to distribute. This setting coincides with classic fair division problems, which allows us to apply concepts from that literature in parallel to equalized odds. In particular, we consider the axioms of resource monotonicity, consistency, and population monotonicity, all three of which relate different allocation instances to prevent paradoxes. Using a geometric characterization of equalized odds, we examine the compatibility of equalized odds with these axioms. We empirically evaluate the cost of allocation rules that satisfy both equalized odds and axioms of fair division on a dataset of FICO credit scores.
Author Information
Paul Gölz (Carnegie Mellon University)
Anson Kahng (Carnegie Mellon University)
Ariel Procaccia (Carnegie Mellon University)
Related Events (a corresponding poster, oral, or spotlight)
-
2019 Poster: Paradoxes in Fair Machine Learning »
Thu. Dec 12th 06:45 -- 08:45 PM Room East Exhibition Hall B + C #83
More from the Same Authors
-
2021 Poster: Fair Sortition Made Transparent »
Bailey Flanigan · Gregory Kehne · Ariel Procaccia -
2020 Poster: Neutralizing Self-Selection Bias in Sampling for Sortition »
Bailey Flanigan · Paul Gölz · Anupam Gupta · Ariel Procaccia -
2019 : Putting Ethical AI to the Vote »
Ariel Procaccia -
2019 Oral: Efficient and Thrifty Voting by Any Means Necessary »
Debmalya Mandal · Ariel Procaccia · Nisarg Shah · David Woodruff -
2017 Poster: Collaborative PAC Learning »
Avrim Blum · Nika Haghtalab · Ariel Procaccia · Mingda Qiao -
2015 Poster: Is Approval Voting Optimal Given Approval Votes? »
Ariel Procaccia · Nisarg Shah -
2014 Poster: Diverse Randomized Agents Vote to Win »
Albert Jiang · Leandro Soriano Marcolino · Ariel Procaccia · Tuomas Sandholm · Nisarg Shah · Milind Tambe -
2014 Poster: Learning Optimal Commitment to Overcome Insecurity »
Avrim Blum · Nika Haghtalab · Ariel Procaccia