PoPS: Policy Pruning and Shrinking for Deep Reinforcement Learning
2020·Arxiv