Continuous Control With Ensemble Deep Deterministic Policy Gradients

TL;DR

This paper empirically investigates various components of deep reinforcement learning in continuous control, revealing insights that lead to the development of the ED2 method, which achieves state-of-the-art results with practical simplicity.

Contribution

The paper introduces ED2, a novel ensemble-based approach that combines multiple insights to improve continuous control performance in deep RL.

Findings

Ensembling multiple actors improves performance.

Existing methods are unstable across training conditions.

Posterior sampling exploration outperforms UCB-based methods.

Abstract

The growth of deep reinforcement learning (RL) has brought multiple exciting tools and methods to the field. This rapid expansion makes it important to understand the interplay between individual elements of the RL toolbox. We approach this task from an empirical perspective by conducting a study in the continuous control setting. We present multiple insights of fundamental nature, including: an average of multiple actors trained from the same data boosts performance; the existing methods are unstable across training runs, epochs of training, and evaluation runs; a commonly used additive action noise is not required for effective training; a strategy based on posterior sampling explores better than the approximated UCB combined with the weighted Bellman backup; the weighted Bellman backup alone cannot replace the clipped double Q-Learning; the critics' initialization plays the major role in ensemble-based actor-critic exploration. As a conclusion, we show how existing tools can be brought together in a novel way, giving rise to the Ensemble Deep Deterministic Policy Gradients (ED2) method, to yield state-of-the-art results on continuous control tasks from OpenAI Gym MuJoCo. From the practical side, ED2 is conceptually straightforward, easy to code, and does not require knowledge outside of the existing RL toolbox.