Pseudorehearsal in actor-critic agents with neural network function approximation
This addresses forgetting in reinforcement learning agents, but appears incremental as it applies known pseudorehearsal techniques to a specific agent setup.
The study tackled catastrophic forgetting in reinforcement learning by investigating pseudorehearsal in actor-critic agents with neural networks, finding that it can assist learning and decrease forgetting in a pole balancing task.
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.