Active Fairness in Algorithmic Decision Making
© 2019 Copyright held by the owner/author(s). Society increasingly relies on machine learning models for automated decision making. Yet, efficiency gains from automation have come paired with concern for algorithmic discrimination that can systematize inequality. Recent work has proposed optimal pos...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Association for Computing Machinery (ACM)
2021
|
| Online Access: | https://hdl.handle.net/1721.1/137087 |
| _version_ | 1811077605392121856 |
|---|---|
| author | Noriega-Campero, Alejandro Bakker, Michiel A Garcia-Bulle, Bernardo Pentland, Alex 'Sandy' |
| author2 | Massachusetts Institute of Technology. Media Laboratory |
| author_facet | Massachusetts Institute of Technology. Media Laboratory Noriega-Campero, Alejandro Bakker, Michiel A Garcia-Bulle, Bernardo Pentland, Alex 'Sandy' |
| author_sort | Noriega-Campero, Alejandro |
| collection | MIT |
| description | © 2019 Copyright held by the owner/author(s). Society increasingly relies on machine learning models for automated decision making. Yet, efficiency gains from automation have come paired with concern for algorithmic discrimination that can systematize inequality. Recent work has proposed optimal post-processing methods that randomize classification decisions for a fraction of individuals, in order to achieve fairness measures related to parity in errors and calibration. These methods, however, have raised concern due to the information inefficiency, intra-group unfairness, and Pareto sub-optimality they entail. The present work proposes an alternative active framework for fair classification, where, in deployment, a decision-maker adaptively acquires information according to the needs of different groups or individuals, towards balancing disparities in classification performance. We propose two such methods, where information collection is adapted to group- and individual-level needs respectively. We show on real-world datasets that these can achieve: 1) calibration and single error parity (e.g., equal opportunity); and 2) parity in both false positive and false negative rates (i.e., equal odds). Moreover, we show that by leveraging their additional degree of freedom, active approaches can substantially outperform randomization-based classifiers previously considered optimal, while avoiding limitations such as intra-group unfairness. |
| first_indexed | 2024-09-23T10:45:40Z |
| format | Article |
| id | mit-1721.1/137087 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T10:45:40Z |
| publishDate | 2021 |
| publisher | Association for Computing Machinery (ACM) |
| record_format | dspace |
| spelling | mit-1721.1/1370872023-02-08T19:13:20Z Active Fairness in Algorithmic Decision Making Noriega-Campero, Alejandro Bakker, Michiel A Garcia-Bulle, Bernardo Pentland, Alex 'Sandy' Massachusetts Institute of Technology. Media Laboratory © 2019 Copyright held by the owner/author(s). Society increasingly relies on machine learning models for automated decision making. Yet, efficiency gains from automation have come paired with concern for algorithmic discrimination that can systematize inequality. Recent work has proposed optimal post-processing methods that randomize classification decisions for a fraction of individuals, in order to achieve fairness measures related to parity in errors and calibration. These methods, however, have raised concern due to the information inefficiency, intra-group unfairness, and Pareto sub-optimality they entail. The present work proposes an alternative active framework for fair classification, where, in deployment, a decision-maker adaptively acquires information according to the needs of different groups or individuals, towards balancing disparities in classification performance. We propose two such methods, where information collection is adapted to group- and individual-level needs respectively. We show on real-world datasets that these can achieve: 1) calibration and single error parity (e.g., equal opportunity); and 2) parity in both false positive and false negative rates (i.e., equal odds). Moreover, we show that by leveraging their additional degree of freedom, active approaches can substantially outperform randomization-based classifiers previously considered optimal, while avoiding limitations such as intra-group unfairness. 2021-11-02T14:16:36Z 2021-11-02T14:16:36Z 2019 2021-06-30T18:25:23Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/137087 Noriega-Campero, Alejandro, Bakker, Michiel A, Garcia-Bulle, Bernardo and Pentland, Alex 'Sandy'. 2019. "Active Fairness in Algorithmic Decision Making." AIES 2019 - Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society. en 10.1145/3306618.3314277 AIES 2019 - Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Association for Computing Machinery (ACM) ACM |
| spellingShingle | Noriega-Campero, Alejandro Bakker, Michiel A Garcia-Bulle, Bernardo Pentland, Alex 'Sandy' Active Fairness in Algorithmic Decision Making |
| title | Active Fairness in Algorithmic Decision Making |
| title_full | Active Fairness in Algorithmic Decision Making |
| title_fullStr | Active Fairness in Algorithmic Decision Making |
| title_full_unstemmed | Active Fairness in Algorithmic Decision Making |
| title_short | Active Fairness in Algorithmic Decision Making |
| title_sort | active fairness in algorithmic decision making |
| url | https://hdl.handle.net/1721.1/137087 |
| work_keys_str_mv | AT noriegacamperoalejandro activefairnessinalgorithmicdecisionmaking AT bakkermichiela activefairnessinalgorithmicdecisionmaking AT garciabullebernardo activefairnessinalgorithmicdecisionmaking AT pentlandalexsandy activefairnessinalgorithmicdecisionmaking |