mlsanityData to 7 Oct 2026

Reinforcement Learning with Stepwise Fairness Constraints

arXiv 2211.03994 ↗First posted 8 Nov 2022cs.LG cs.AI cs.CYcited by 9

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 26 of its 39 references that are in the map. First posted on arXiv 8 Nov 2022.

Builds on

26 of 39 references.

References · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2017Mask R-CNNcited by 7,259
  3. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  4. 2017Counterfactual Fairnesscited by 802
  5. 2018A Reductions Approach to Fair Classificationcited by 585
  6. 2018Fairness Without Demographics in Repeated Loss Minimizationcited by 363
  7. 2015Strategic Classificationcited by 256
  8. 2018Delayed Impact of Fair Machine Learningcited by 183
  9. 2016Fairness in Learning: Classic and Contextual Banditscited by 167
  10. 2016Fairness in Reinforcement Learningcited by 93
  11. 2020How Do Fair Decisions Fare in Long-term Qualification?cited by 59
  12. 2021When and How Mixup Improves Calibrationcited by 59
14 more
  1. 2019Algorithms for Fairness in Sequential Decision Makingcited by 53
  2. 2019Causal Modeling for Fairness in Dynamical Systemscited by 46
  3. 2020Toward Better Generalization Bounds with Locally Elastic Stabilitycited by 41
  4. 2021Adversarial Training Helps Transfer Learning via Better Representationscited by 36
  5. 2020Constrained episodic reinforcement learning in concave-convex and knapsack settingscited by 33
  6. 2020Representation via Representations: Domain Generalization via Adversarially Learned Invariant Representationscited by 26
  7. 2020Improving Adversarial Robustness via Unlabeled Out-of-Domain Datacited by 23
  8. 2022FIFA: Making Fairness More Generalizable in Classifiers Trained on Imbalanced Datacited by 13
  9. 2022Socially Fair Reinforcement Learningcited by 13
  10. 2020Decision-Aware Conditional GANs for Time Series Datacited by 11
  11. 2021Understanding Dynamics of Nonlinear Representation Learning and Its Applicationcited by 11
  12. 2020Interpreting Robust Optimization via Adversarial Influence Functionscited by 8
  13. 2019Extracting Incentives from Black-Box Decisionscited by 7
  14. 2020Towards Understanding the Dynamics of the First-Order Adversariescited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 7; 2024: 0; 2025: 0; 2026: 2.

9 papers.

  1. 2023Datasets and Benchmarks for Offline Safe Reinforcement Learningcited by 58
  2. 2023Last-Layer Fairness Fine-tuning is Simple and Effective for Neural Networkscited by 17
  3. 2023Achieving Fairness in Multi-Agent Markov Decision Processes Using Reinforcement Learningcited by 8
  4. 2023Multi-objective Reinforcement Learning with Nonlinear Preferences: Provable Approximation for Maximizing Expected Scalarized Returncited by 7
  5. 2023Remembering to Be Fair: Non-Markovian Fairness in Sequential Decision Makingcited by 7
  6. 2023Multi-Agent Reinforcement Learning: Methods, Applications, Visionary Prospects, and Challengescited by 5
  7. 2023Fair Supervised Learning with A Simple Random Sampler of Sensitive Attributescited by 2
  8. 2026Price of Fairness in Short-Term and Long-Term Algorithmic Selectionscited by 0
  9. 2026Simulating Eutopia: Revisiting Long-term Fairness with Outcomes, Performativity, and Dynamicscited by 0