Trust region methods, such as TRPO, are often used to stabilize policy optimization algorithms in reinforcement learning (RL). While current trust region strategies are effective for continuous control, they typically require a prohibitively large amount of on-policy interaction with the environment. To address this problem, we propose an off-policy trust region method, Trust-PCL. The algorithm is the result of observing that the optimal policy and state values of a maximum reward objective with a relative-entropy regularizer satisfy a set of multi-step pathwise consistencies along any path. Thus, Trust-PCL is able to maintain optimization stability while exploiting off-policy data to improve sample efficiency. When evaluated on a number of continuous control tasks, Trust-PCL improves the solution quality and sample efficiency of TRPO.
@article{arxiv.1707.01891,
title = {Trust-PCL: An Off-Policy Trust Region Method for Continuous Control},
author = {Ofir Nachum and Mohammad Norouzi and Kelvin Xu and Dale Schuurmans},
journal= {arXiv preprint arXiv:1707.01891},
year = {2018}
}