Reward Centering

We show that discounted methods for solving continuing reinforcement learning problems can perform significantly better if they center their rewards by subtracting out the rewards' empirical average. The improvement is substantial at commonly used discount factors and increases further as the discount factor approaches one. In addition, we show that if a _problem's_ rewards are shifted by a constant, then standard methods perform much worse, whereas methods with reward centering are unaffected. Estimating the average reward is straightforward in the on-policy setting; we propose a slightly more sophisticated method for the off-policy setting. Reward centering is a general idea, so we expect almost every reinforcement-learning algorithm to benefit by the addition of reward centering.

Authors

• Abhishek Naik
• Yi Wan School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China Technology Innovation Center for Collaborative Applications of Natural Resources Data in GBA, Ministry of Natural Resources, Guangzhou 510075, China Hubei LuoJia Laboratory, Wuhan 430079, China
• Manan Tomar
• Richard S. Sutton DeepMind, Edmonton, Alberta, Canada University of Alberta, Edmonton, Alberta, Canada Alberta Machine Intelligence Institute (Amii), Edmonton, Alberta, Canada Canada CIFAR AI Chair, Canada

Keywords

No keywords are indexed for this paper.