Publications: Reinforcement Learning
Reinforcement Learning tasks are learning problems where the desired behavior is not known; only sparse feedback on how well the agent is doing is provided. Reinforcement Learning techniques include value-function and policy iteration methods (note that although evolutionary computation and neuroevolution can also be seen as reinforcement learning methods, they are presented separately in this area hierarchy.)
- Using Active Relocation to Aid Reinforcement Learning
[Details] [PDF]
Lilyana Mihalkova and Raymond Mooney
In Prodeedings of the 19th International FLAIRS Conference (FLAIRS-2006), 580-585, Melbourne Beach, FL, May 2006.We propose a new framework for aiding a reinforcement learner by allowing it to relocate, or move, to a state it selects so as to decrease the number of steps it needs to take in order to develop an effective policy. The framework requires a minimal amount of human involvement or expertise and assumes a cost for each relocation. Several methods for taking advantage of the ability to relocate are proposed, and their effectiveness is tested in two commonly-used domains.
ML ID: 166
- Guiding a Reinforcement Learner with Natural Language Advice: Initial Results in RoboCup Soccer
[Details] [PDF]
Gregory Kuhlmann, Peter Stone, Raymond J. Mooney, and Jude W. Shavlik
In The AAAI-2004 Workshop on Supervisory Control of Learning and Adaptive Systems, July 2004.We describe our current efforts towards creating a reinforcement learner that learns both from reinforcements provided by its environment and from human-generated advice. Our research involves two complementary components: (a) mapping advice expressed in English to a formal advice language and (b) using advice expressed in a formal notation in a reinforcement learner. We use a subtask of the challenging RoboCup simulated soccer task as our testbed.
ML ID: 151
