Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning

arXiv Preprint 2019

Xue Bin Peng    Aviral Kumar    Grace Zhang    Sergey Levine

University of California, Berkeley


In this paper, we aim to develop a simple and scalable reinforcement learning algorithm that uses standard supervised learning methods as subroutines. Our goal is an algorithm that utilizes only simple and convergent maximum likelihood loss functions, while also being able to leverage off-policy data. Our proposed approach, which we refer to as advantage-weighted regression (AWR), consists of two standard supervised learning steps: one to regress onto target values for a value function, and another to regress onto weighted target actions for the policy. The method is simple and general, can accommodate continuous and discrete actions, and can be implemented in just a few lines of code on top of standard supervised learning methods. We provide a theoretical motivation for AWR and analyze its properties when incorporating off-policy data from experience replay. We evaluate AWR on a suite of standard OpenAI Gym benchmark tasks, and show that it achieves competitive performance compared to a number of well-established state-of-the-art RL algorithms. AWR is also able to acquire more effective policies than most off-policy algorithms when learning from purely static datasets with no additional environmental interactions. Furthermore, we demonstrate our algorithm on challenging continuous control tasks with highly complex simulated characters.

Paper: [PDF]       Code: [GitHub]       Preprint: [arXiv]



	author = {Xue Bin Peng and Aviral Kumar and Grace Zhang and Sergey Levine},
	title = {Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning},
	journal = {CoRR},
	volume = {abs/1910.00177},
	year = {2019},
	url = {},
	archivePrefix = {arXiv},
	eprint = {1910.00177},
	timestamp = {Tue, 01 October 2019 11:27:50 +0200},
	bibsource = {dblp computer science bibliography,}