Combining Imitation and Reinforcement Learning with Free Energy Principle
Imitation Learning (IL) and Reinforcement Learning (RL) from high dimensional sensory inputs are often introduced as separate problems, but a more realistic problem setting is how to merge the techniques so that the agent can reduce exploration costs by partially imitating experts at the same time it maximizes its return. Even when the experts are suboptimal (e.g. Experts learned halfway with other RL methods or human-crafted experts), it is expected that the agent outperforms the suboptimal experts’ performance. In this paper, we propose to address the issue by using and theoretically extending Free Energy Principle, a unified brain theory that explains perception, action and model learning in a Bayesian probabilistic way. Our results show that our approach achieves at least equivalent or better performance than standard IL or RL in visual control tasks with sparse rewards.
PDF Abstract
DeepMind Control Suite
