Improving Reinforcement Learning with Human Input

Improving Reinforcement Learning with Human Input

Matthew E. Taylor

Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence
Early Career. Pages 5724-5728. https://doi.org/10.24963/ijcai.2018/817

Reinforcement learning (RL) has had many successes when learning autonomously. This paper and accompanying talk consider how to make use of a non-technical human participant, when available. In particular, we consider the case where a human could 1) provide demonstrations of good behavior, 2) provide online evaluative feedback, or 3) define a curriculum of tasks for the agent to learn on. In all cases, our work has shown such information can be effectively leveraged. After giving a high-level overview of this work, we will highlight a set of open questions and suggest where future work could be usefully focused.
Keywords:
Machine Learning: Reinforcement Learning
Humans and AI: Human-Computer Interaction
Machine Learning: Transfer, Adaptation, Multi-task Learning