Reading Group on Statistical Learning, Fall 2003

Computer Science Department and Statistics Department, UC Irvine

Organizers: Padhraic Smyth (CS) and David Van Dyk (Statistics)

Logistics: Meet on alternate Tuesdays from 4 to 5pm in CS 432

Note that on the other Tuesdays we have the AI and Statistics seminar series.

September 30th:  Topic Models for Documents
Latent Dirichlet Allocation  
David M. Blei, Andrew Y. Ng, Michael I. Jordan, Journal of Machine Learning Research, 3(Jan):993-1022, 2003.

Finding Scientific Topics.
Griffiths, T., & Steyvers, M. (submitted), Proceedings of the National Academy of Sciences.

Discussant: Michal Rosen-Zvi (michal@ics.uci.edu)

October 14th: Conditional  Random Fields
Efficiently Inducing Features of Conditional Random Fields .
Andrew McCallum, Uncertainty in Artificial Intelligence Conference (UAI), 2003.

Shallow Parsing with Conditional Random Fields
 F. Sha and F. Pereira, Proceedings of Human Language Technology, NAACL, 2003

Discussant: David Van Dyk (dvd@ics.uci.edu)

October 28th: Non-Parametric Priors and Tree-Structured Priors

Defining Priors for Distributions using Dirichlet Diffusion Trees.
Radford Neal, Technical Report, 2001.

Hierarchical Topic Models and the Nested Chinese Restaurant Process.
Blei et al, NIPS 2003, to appear.


Discussant: Max Welling (welling@ics.uci.edu)

 



November 11th: "Information Bottleneck" algorithms
The Information Bottleneck EM Algorithm.
Elidan and Friedman, Uncertainty in Artificial Intelligence Conference (UAI), 2003.

Related Paper:
Maximum Likelihood and the Information Bottleneck
Slonim and Weiss, NIPS 2002.

Discussant: Se Young Kim (sykim@ics.uci.edu)




BELOW ARE SUGGESTED TOPICS/PAPERS FOR THE FUTURE:
THESE ARE SUBJECT TO CHANGE DEPENDING ON THE GROUP'S INTERESTS.


November ?: SVMs and Kernel-based Learning
An Introduction to Kernel-Based Learning Algorithms
K. Muller et al., IEEE Transactions on Neural Networks, 2001

Max-Margin Markov Networks
B. Taskar, C. Guestrin, and D. Koller, NIPS Conference, to appear, December 2003.

Discussant: TBD

November ?: Probabilistic Learning and Inference in Computer Vision

Bayesian Object Localisation in Images
O Sullivan, Blake, Isard, and McCormick, IJCCV, 2001.

Real-Valued Graphical Models for Computer Vision
Michael Isard, CVPR, 2003.

Discussant: TBD



Other Possible Topics (TBD) - for November 11th and 25th