Reward Distribution

Distributional reinforcement learning (DRL) is a paradigm in machine learning that focuses on predicting the distribution of potential future rewards rather than a single expected value.