• UCL Twitter account
  • UCL YouTube channel
  • UCL Facebook page
  • UCL SoundCloud channel
  • UCL iTunes store

Information for Staff

Calendar

Select dates to view past and future events

16:00 - 17:00 28 September 2011

Optimal Reinforcement Learning for Gaussian Systems

Location

Seminar Room B10 (Basement) | Alexandra House (link Map)
17-19 Queen square | London | WC1N 3AR | United Kingdom

Open to: Academic | Public | Student

Speaker information

Philipp Hennig

Host: Gatsby Unit External Seminar Series The exploration-exploitation tradeoff is among the central challenges of reinforcement learning. The exact Bayesian learner providing the optimal solution is intractable in general. Like in other inference tasks, though, convenient prior assumptions can allow exact statements. In this talk, I will show that, in the case of Gaussian process inference, it is possible to make analytic statements about optimal learning of both loss function and transition dynamics of nonlinear, time-varying systems in continuous time and space, subject to a relatively weak restriction on the dynamics. The main result is a theoretical insight, but I will also show some approximate numerical results.


Contact

Gatsby Unit
44 (0)20 7679 1176 |