Program

All talks took place on the campus of McGill University in downtown Montreal. See the venue page for maps and details.

Thursday, June 18
(in Rutherford 112)

12:30-1:30 Poster Setup Time

Welcome Statement by Doina Precup, McGill, Canada (Video)

1:30 Rich Sutton, University of Alberta, Canada
       Deconstructing Reinforcement Learning (Video)

2:30 Terrence Sejnowski, The Salk Institute and the University of California, San Diego, USA
       Birdsong Learning (Video)

3:30 30-min Break

4:00 Gerald Tesauro, IBM Research, USA
       Fifty Years of RL in Games (Video)

5:00 Andrew Ng, Stanford University, USA
        Reinforcement Learning, Apprenticeship Learning, and Robotic Control (Video)

6:00-6:20 David Wingate, MIT
         Report from 2009 RL Competition
(Video)


6:30 Poster Dinner (in Arts/Leacock Lobby)


Friday, June 19 (in McConnell Engineering 304)

8:50 Dimitri Bertsekas, Massachusetts Institute of Technology, USA
       On Temporal Difference Methods and Extensions (Video)

9:50 Wolfram Schultz, University of Cambridge, UK
       Neural Reward Signals Complying with Theoretical Constructs of Contiguity,
       Contingency and Prediction Error

10:50 30-min Break

11:20 Warren B. Powell, Princeton University, USA
        
Approximate Dynamic Programming: Solving the curses of dimensionality (Video)

12:20-2:00 Lunch

2:00 Peter Dayan, University College London, UK
        Three and a Half Theories of Neural Reinforcement Learning (Video)

3:00 15-minute Break

3:15 Ben Van Roy, Stanford University, USA
       Reflections on Temporal Difference Learning (Video)

4:15 Andrew Barto, University of Massachusetts, Amherst, USA
       
Beyond Reinforcement Learning (Video)

5:15 End of MSRL

(some videos can also be downloaded from here)
Comments