Novelty-based Sample Reuse for Continuous Robotics Control

Ke Duan, Kai Yang, Houde Liu +1

In reinforcement learning, agents collect state information and rewards through environmental interactions, essential for policy refinement. This process is notably time-consuming,…