Recursive Reward Aggregation
This work addresses the problem of flexible behavior alignment in RL for practitioners, though it appears incremental as it builds on existing MDP frameworks with new aggregation methods.
The paper tackles the challenge of aligning agent behavior with complex objectives in reinforcement learning by proposing an alternative approach that uses recursive reward aggregation functions instead of modifying the reward function, showing that it effectively optimizes diverse objectives in experiments.
In reinforcement learning (RL), aligning agent behavior with specific objectives typically requires careful design of the reward function, which can be challenging when the desired objectives are complex. In this work, we propose an alternative approach for flexible behavior alignment that eliminates the need to modify the reward function by selecting appropriate reward aggregation functions. By introducing an algebraic perspective on Markov decision processes (MDPs), we show that the Bellman equations naturally emerge from the recursive generation and aggregation of rewards, allowing for the generalization of the standard discounted sum to other recursive aggregations, such as discounted max and Sharpe ratio. Our approach applies to both deterministic and stochastic settings and integrates seamlessly with value-based and actor-critic algorithms. Experimental results demonstrate that our approach effectively optimizes diverse objectives, highlighting its versatility and potential for real-world applications.