Reinforcement Learning with Stepwise Fairness Constraints

Zhun Deng; He Sun; Zhiwei Steven Wu; Linjun Zhang; David C. Parkes

Reinforcement Learning with Stepwise Fairness Constraints

Machine Learning 2022-11-09 v1 Artificial Intelligence Computers and Society

Authors: Zhun Deng , He Sun , Zhiwei Steven Wu , Linjun Zhang , David C. Parkes

Abstract

AI methods are used in societally important settings, ranging from credit to employment to housing, and it is crucial to provide fairness in regard to algorithmic decision making. Moreover, many settings are dynamic, with populations responding to sequential decision policies. We introduce the study of reinforcement learning (RL) with stepwise fairness constraints, requiring group fairness at each time step. Our focus is on tabular episodic RL, and we provide learning algorithms with strong theoretical guarantees in regard to policy optimality and fairness violation. Our framework provides useful tools to study the impact of fairness constraints in sequential settings and brings up new challenges in RL.

Keywords

reinforcement learning reinforcement learning from human feedback fairness in machine learning

Cite

@article{arxiv.2211.03994,
  title  = {Reinforcement Learning with Stepwise Fairness Constraints},
  author = {Zhun Deng and He Sun and Zhiwei Steven Wu and Linjun Zhang and David C. Parkes},
  journal= {arXiv preprint arXiv:2211.03994},
  year   = {2022}
}

Comments

Fairness, Reinforcement Learning

Reinforcement Learning with Stepwise Fairness Constraints

Abstract

Keywords

Cite

Comments

Related papers