Adapting Behaviour for Learning Progress
This addresses the problem of tuning exploration for reinforcement learning agents, offering a more efficient solution for researchers and practitioners, though it appears incremental as it builds on existing distributed agent frameworks.
The paper tackles the challenge of exploration in reinforcement learning by dynamically adapting data generation using a non-stationary multi-armed bandit to optimize learning progress, demonstrating results comparable to per-task tuning on Atari 2600 games at a fraction of the cost.
Determining what experience to generate to best facilitate learning (i.e. exploration) is one of the distinguishing features and open challenges in reinforcement learning. The advent of distributed agents that interact with parallel instances of the environment has enabled larger scales and greater flexibility, but has not removed the need to tune exploration to the task, because the ideal data for the learning algorithm necessarily depends on its process of learning. We propose to dynamically adapt the data generation by using a non-stationary multi-armed bandit to optimize a proxy of the learning progress. The data distribution is controlled by modulating multiple parameters of the policy (such as stochasticity, consistency or optimism) without significant overhead. The adaptation speed of the bandit can be increased by exploiting the factored modulation structure. We demonstrate on a suite of Atari 2600 games how this unified approach produces results comparable to per-task tuning at a fraction of the cost.