The Max-Min Formulation of Multi-Objective Reinforcement Learning: From Theory to a Model-Free Algorithm
This addresses fairness in multi-objective reinforcement learning for real-world applications, but it appears incremental as it builds on existing frameworks.
The paper tackles multi-objective reinforcement learning by introducing a max-min framework for fairness among goals, resulting in a model-free algorithm that shows notable performance improvements over baselines.
In this paper, we consider multi-objective reinforcement learning, which arises in many real-world problems with multiple optimization goals. We approach the problem with a max-min framework focusing on fairness among the multiple goals and develop a relevant theory and a practical model-free algorithm under the max-min framework. The developed theory provides a theoretical advance in multi-objective reinforcement learning, and the proposed algorithm demonstrates a notable performance improvement over existing baseline methods.