Repository logo
 

Interpolated policy gradient: Merging on-policy and off-policy gradient estimation for deep reinforcement learning

Accepted version
Peer-reviewed

Type

Conference Object

Change log

Authors

Gu, S 
Lillicrap, T 
Turner, RE 
Schölkopf, B 

Abstract

Off-policy model-free deep reinforcement learning methods using previously collected data can improve sample efficiency over on-policy policy gradient techniques. On the other hand, on-policy algorithms are often more stable and easier to use. This paper examines, both theoretically and empirically, approaches to merging on- and off-policy updates for deep reinforcement learning. Theoretical results show that off-policy updates with a value function estimator can be interpolated with on-policy policy gradient updates whilst still satisfying performance bounds. Our analysis uses control variate methods to produce a family of policy gradient algorithms, with several recently proposed algorithms being special cases of this family. We then provide an empirical comparison of these techniques with the remaining algorithmic details fixed, and show how different mixing of off-policy gradient estimates with on-policy samples contribute to improvements in empirical performance. The final algorithm provides a generalization and unification of existing deep policy gradient techniques, has theoretical guarantees on the bias introduced by off-policy updates, and improves on the state-of-the-art model-free deep RL methods on a number of OpenAI Gym continuous control benchmarks.

Description

Keywords

cs.LG, cs.LG, cs.AI, cs.RO

Journal Title

Advances in Neural Information Processing Systems

Conference Name

Journal ISSN

1049-5258

Volume Title

2017-December

Publisher

Sponsorship
Engineering and Physical Sciences Research Council (EP/J012300/1)
EPSRC (via University of Sheffield) (143103)