Improving Interactive Reinforcement Agent Planning with Human Demonstration
This is an incremental improvement for interactive reinforcement learning agents, specifically targeting cost reduction in human-robot teaching scenarios.
The paper tackles the high learning cost of TAMER agents using UCT planning by initializing the reward function with inverse reinforcement learning from human demonstrations, resulting in improved learning efficiency with reduced feedback and incorrect actions, and faster convergence to an optimal policy.
TAMER has proven to be a powerful interactive reinforcement learning method for allowing ordinary people to teach and personalize autonomous agents' behavior by providing evaluative feedback. However, a TAMER agent planning with UCT---a Monte Carlo Tree Search strategy, can only update states along its path and might induce high learning cost especially for a physical robot. In this paper, we propose to drive the agent's exploration along the optimal path and reduce the learning cost by initializing the agent's reward function via inverse reinforcement learning from demonstration. We test our proposed method in the RL benchmark domain---Grid World---with different discounts on human reward. Our results show that learning from demonstration can allow a TAMER agent to learn a roughly optimal policy up to the deepest search and encourage the agent to explore along the optimal path. In addition, we find that learning from demonstration can improve the learning efficiency by reducing total feedback, the number of incorrect actions and increasing the ratio of correct actions to obtain an optimal policy, allowing a TAMER agent to converge faster.