Reinforcement learning is a commonly used technique for optimising objectives in decision support systems for complex problem solving. When these systems affect individuals or groups, it is essential to reflect on fairness. As absolute fairness is in practice not achievable, we propose a framework which allows to balance distinct fairness notions along with the primary objective. To this end, we formulate group and individual fairness in sequential fairness notions. First, we present an extended Markov decision process, f MDP, that is explicitly aware of individuals and groups. Next, we formalise fairness notions in terms of this f MDP which allows us to evaluate the primary objective along with the fairness notions that are important to the user, taking a multi-objective reinforcement learning approach. To evaluate our framework, we consider two scenarios that require distinct aspects of the performance-fairness trade-off: job hiring and fraud detection. The objectives in job hiring are to compose strong teams, while providing equal treatment to similar individual applicants and to groups in society. The trade-off in fraud detection is the necessity of detecting fraudulent transactions, while distributing the burden for customers of checking transactions fairly. In this framework, we further explore the influence of distance metrics on individual fairness and highlight the impact of the history size on the fairness calculations and the obtainable fairness through exploration.
Cimpean, IA, Jonker, CM, Libin, P & Nowe, A 2024, A Reinforcement Learning Framework For Studying Group And Individual Fairness: Extended Abstract. in The 23rd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2024: Extended Abstract. Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS, International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), pp. 2216-2218, 23rd International Conference on Autonomous Agents and Multi-Agent Systems - AAMAS 2024, Auckland, New Zealand, 6/05/24. <https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p2216.pdf>
Cimpean, I. A., Jonker, C. M., Libin, P., & Nowe, A. (2024). A Reinforcement Learning Framework For Studying Group And Individual Fairness: Extended Abstract. In The 23rd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2024: Extended Abstract (pp. 2216-2218). (Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS). International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS). https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p2216.pdf
@inproceedings{fb19678139db488a98cbb60eef7728e1,
title = "A Reinforcement Learning Framework For Studying Group And Individual Fairness: Extended Abstract",
abstract = "Reinforcement learning is a commonly used technique for optimising objectives in decision support systems for complex problem solving. When these systems affect individuals or groups, it is essential to reflect on fairness. As absolute fairness is in practice not achievable, we propose a framework which allows to balance distinct fairness notions along with the primary objective. To this end, we formulate group and individual fairness in sequential fairness notions. First, we present an extended Markov decision process, f MDP, that is explicitly aware of individuals and groups. Next, we formalise fairness notions in terms of this f MDP which allows us to evaluate the primary objective along with the fairness notions that are important to the user, taking a multi-objective reinforcement learning approach. To evaluate our framework, we consider two scenarios that require distinct aspects of the performance-fairness trade-off: job hiring and fraud detection. The objectives in job hiring are to compose strong teams, while providing equal treatment to similar individual applicants and to groups in society. The trade-off in fraud detection is the necessity of detecting fraudulent transactions, while distributing the burden for customers of checking transactions fairly. In this framework, we further explore the influence of distance metrics on individual fairness and highlight the impact of the history size on the fairness calculations and the obtainable fairness through exploration.",
author = "Cimpean, {Ioana Alexandra} and Jonker, {Catholijn M} and Pieter Libin and Ann Nowe",
note = "Funding Information: Alexandra Cimpean receives funding from the Fonds voor Wetenschappelijk Onderzoek (FWO) via fellowship grant 1SF7823N. All experiments were performed on the VSC high performance computing infrastructure [1]. Publisher Copyright: {\textcopyright} 2024 International Foundation for Autonomous Agents and Multiagent Systems.; 23rd International Conference on Autonomous Agents and Multi-Agent Systems - AAMAS 2024, AAMAS 2024 ; Conference date: 06-05-2024 Through 10-05-2024",
year = "2024",
language = "English",
series = "Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS",
publisher = "International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)",
pages = "2216--2218",
booktitle = "The 23rd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2024",
url = "https://www.aamas2024-conference.auckland.ac.nz/",
}