Reinforcement learning (RL) is an effective framework for online learning by autonomous agents. Most RL research focuses on domain-independent learning emphalgorithms, requiring an expert human to define the emphenvironment (state and action representation) and emphtask to be performed (e.g. start state and reward function) on a case-by-case basis. In this paper, we describe a general language for a teacher to specify sequential decision making tasks to RL agents. The teacher may communicate properties such as start states, reward functions, termination conditions, successful execution traces, task decompositions, and other advice. The learner may then practice and learn the task on its own using any RL algorithm. We demonstrate our language in a simple GridWorld example and on the RoboCup soccer keepaway benchmark problem. The language forms the basis of a larger ``Bootstrap Learning'' model for machine learning, a paradigm for incremental development of complete systems through integration of multiple machine learning techniques.

PDF, PS, HTML

In AAAI Spring 2009 Symposium on Agents that Learn from Human Teachers, March 2008.

@InProceedings{AAAIsymp09-fasel, title={A General Purpose Task Specification Language for Bootstrap Learning}, author={Ian Fasel and Michael Quinlan and Peter Stone}, booktitle={AAAI Spring 2009 Symposium on Agents that Learn from Human Teachers}, month={March}, url="http://www.cs.utexas.edu/users/ai-lab?AAAIsymp09-fasel", year={2008} }