Welcome
These pages are here for historical interest. You probably want the latest pages instead.
Our schedule is flexible, but roughly decided a few weeks in advance by those that come along. Each week you should try to read at least one of the papers suggested.
Please take a look at our list of possible future topics, which also links to supplementary reading.
Rough Schedule
- Future -- possible ideas
- 28th November 2003 - Boosting, Maximum Likelihood and Bregman divergence
- 21st November 2003 - CFG
- 14th November 2003 - Active Learning
- 7th November 2003 - Exact Sampling
- 31st October 2003 - Gaussian processes for density estimation
- 24th October 2003 - Ranking on Data Manifolds
- Friday 10th October 2003 - Cross Entropy & EM for RL
- 15th September 2003 - Dirichlet Diffusion Trees
- 8th September 2003 - Genetic algorithms
- 1st September 2003 - Dirichlet Diffusion Trees
- Thursday 28th August 2003 - Stick-breaking priors
- 11th August 2003 - Dirichlet Processes
- 28th July 2003 - GBP in hybrid networks
- 21st July 2003 - Partition Function upper bounds
- 14th July 2003 - Data visualisation
- 7th July 2003 - Survey Propagation
- Friday 27th June 2003 - Bayesian RL?
- Friday 20th June 2003 - LLE
- 2nd June 2003 - No Meeting
- Wednesday 28th May 2003 2pm - Reinforcement Learning
- 19th May 2003 - Digital Fountain Codes
- 12th May 2003 - Reinforcement Learning
- 5th May 2003 - Bank Holiday
- 28th April 2003 - EP for Gaussian Process Classification
- 21st April 2003 - Bank Holiday
- 14th April 2003 - Gaussian Processes (cont)
- 7th April 2003 - Gaussian Processes
- 31st March 2003 - Week off
- Wednesday 26 March 2003 - Expectation Propagation
- 17th March 2003 - loopy BP
- 10th March 2003 - Graphical models
- 3rd March 2003 - semi-definite and second-order cone programming
- 24th February 2003 - SMO and SVMs
- 17th February 2003 - matrix perturbation, quadratic programming
- 10th February 2003 - Game Theory
- 3rd February 2003 - Linear programming
- 27rd January 2003 - Linear Programming
- 20th January 2003 - Spectral methods