Catastrophic forgetting has a significant negative impact in reinforcement learning. The purpose of this study is to investigate how pseudorehearsal can change performance of an actor-critic agent with neural-network function approximation. We tested agent in a pole balancing task and compared different pseudorehearsal approaches. We have found that pseudorehearsal can assist learning and decrease forgetting.
@article{arxiv.1712.07686,
title = {Pseudorehearsal in actor-critic agents with neural network function approximation},
author = {Vladimir Marochko and Leonard Johard and Manuel Mazzara and Luca Longo},
journal= {arXiv preprint arXiv:1712.07686},
year = {2018}
}