Fairness in Reinforcement Learning
This work addresses fairness concerns in RL for applications where algorithmic decisions impact environments and rewards, though it is incremental as it builds on existing RL frameworks.
The paper tackles the problem of fairness in reinforcement learning by introducing a fairness constraint that prohibits preferring an action with lower long-term reward, and shows that exact fairness requires exponential time for non-trivial approximation, while a polynomial-time algorithm is possible under approximate fairness.
We initiate the study of fairness in reinforcement learning, where the actions of a learning algorithm may affect its environment and future rewards. Our fairness constraint requires that an algorithm never prefers one action over another if the long-term (discounted) reward of choosing the latter action is higher. Our first result is negative: despite the fact that fairness is consistent with the optimal policy, any learning algorithm satisfying fairness must take time exponential in the number of states to achieve non-trivial approximation to the optimal policy. We then provide a provably fair polynomial time algorithm under an approximate notion of fairness, thus establishing an exponential gap between exact and approximate fairness