Perceptual Science Series

Efficient Model Learning for Reinforcement Learning

Dr. Michael Littman

Monday, October 22, 2007, 01:00pm - 02:00pm

Rutgers University, Department of Computer Science

Copy to My Calendar (iCal) Download as iCal file
 

This talk addresses the problem of learning efficiently to make
sequential decisions with a particular focus on generalizing
experience without forfeiting formal learning-time guarantees.  I'll
summarize the theoretically motivated algorithms my group has been
developing that exhibit practical advantages over existing learning
algorithms.  I'll also toss in some video footage of robots learning
to move around by exploring efficiently.

Background Reading:

Reinforcement Learning:  A Survey

 

Dr. Michael Littman