Sample Efficient Actor-Critic with Experience Replay

About

This paper presents an actor-critic deep reinforcement learning agent with experience replay that is stable, sample efficient, and performs remarkably well on challenging environments, including the discrete 57-game Atari domain and several continuous control problems. To achieve this, the paper introduces several innovations, including truncated importance sampling with bias correction, stochastic dueling network architectures, and a new trust region policy optimization method.

Ziyu Wang, Victor Bapst, Nicolas Heess, Volodymyr Mnih, Remi Munos, Koray Kavukcuoglu, Nando de Freitas · 2016

Related benchmarks

Task	Dataset	Result
Reinforcement Learning	Supply Chain Optimization Environment (test)	Max Reward17.8	10
Reinforcement Learning	Atari 49 ALE (test)	Alien Score390.2	6
Reinforcement Learning	Atari 49 games	Median Human-Normalized Score46.8	3