Sample-Efficient Deep Reinforcement Learning via Episodic Backward Update

Su Young Lee, Sungik Choi, Sae-Young Chung

Abstract:

We propose Episodic Backward Update (EBU) – a novel deep reinforcement learning algorithm with a direct value propagation. In contrast to the conventional
use of the experience replay with uniform random sampling, our agent samples
a whole episode and successively propagates the value of a state to its previous
states. Our computationally efficient recursive algorithm allows sparse and delayed rewards to propagate directly through all transitions of the sampled episode.
We theoretically prove the convergence of the EBU method and experimentally
demonstrate its performance in both deterministic and stochastic environments.
Especially in 49 games of Atari 2600 domain, EBU achieves the same mean and
median human normalized performance of DQN by using only 5% and 10% of
samples, respectively.

Track: Replicability

NeurIPS Paper Id: https://openreview.net/forum?id=HyleUNBeIr