AAAI Publications, Twenty-Second International Joint Conference on Artificial Intelligence

Font Size: 
Automatic State Abstraction from Demonstration
Luis Carlos Cobo, Peng Zang, Charles Lee Isbell Jr., Andrea Lockerd Thomaz

Last modified: 2011-06-28


Learning from Demonstration (LfD) is a popular technique for building decision-making agents from human help. Traditional LfD methods use demonstrations as training examples for supervised learning, but complex tasks can require more examples than is practical to obtain. We present Abstraction from Demonstration (AfD), a novel form of LfD that uses demonstrations to infer state abstractions and reinforcement learning (RL) methods in those abstract state spaces to build a policy. Empirical results show that AfD is greater than an order of magnitude more sample efficient than jus tusing demonstrations as training examples, and exponentially faster than RL alone.

Full Text: PDF