Only Relevant Information Matters: Filtering Out Noisy Samples To Boost RL

Only Relevant Information Matters: Filtering Out Noisy Samples To Boost RL

Yannis Flet-Berliac, Philippe Preux

Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence
Main track. Pages 2711-2717. https://doi.org/10.24963/ijcai.2020/376

In reinforcement learning, policy gradient algorithms optimize the policy directly and rely on sampling efficiently an environment. Nevertheless, while most sampling procedures are based on direct policy sampling, self-performance measures could be used to improve such sampling prior to each policy update. Following this line of thought, we introduce SAUNA, a method where non-informative transitions are rejected from the gradient update. The level of information is estimated according to the fraction of variance explained by the value function: a measure of the discrepancy between V and the empirical returns. In this work, we use this criterion to select samples that are useful to learn from, and we demonstrate that this selection can significantly improve the performance of policy gradient methods. In this paper: (a) We introduce the SAUNA method to filter transitions. (b) We conduct experiments on a set of benchmark continuous control problems. SAUNA significantly improves performance. (c) We investigate how SAUNA reliably selects samples with the most positive impact on learning and study its improvement on both performance and sample efficiency.
Keywords:
Machine Learning: Deep Reinforcement Learning
Machine Learning: Reinforcement Learning