Reading group: sequence modeling and analysis


Fall 2004
Fri 1:00-2:30 PM, CoRE A (301)

Fri, 9/10/2004: Segmental models: inference and learning

Papers to read:

http://www.ai.mit.edu/~murphyk/Bayes/rabiner.pdf
http://citeseer.ist.psu.edu/mitchell95complexity.html
http://citeseer.ist.psu.edu/90187.html

Notes (by Pai-Hsi Huang).

Fri, 9/17/2004: Maximum entropy models

Papers to read: Stephen Della Pietra, Vincent Della Pietra, John Lafferty, "Inducing Features Of Random Fields," T-PAMI, 1997.
A. Berger, "A brief Maxent tutorial"

Fri, 9/24/2004: Maximum entropy Markov models

Discriminative vs. generative: see a very simple argument on p. 3 (Sec. 3)
Choudhury T., Rehg J. M., Pavlovic V. and Pentland A., Boosting and Structure Learning in Dynamic Bayesian Networks for Audio-Visual Speaker Detection", ICPR, 2002.
Also:
"Bayesian Network Classifiers" N. Friedman, D. Geiger, and M. Goldszmidt, In Machine Learning 29:131--163, 1997.

Papers to read:
Andrew Mccallum, Dayne Freitag, Fernando Pereira, "Maximum Entropy Markov Models for Information Extraction and Segmentation," ICML 2000.
Andrew Mccallum, Dayne Freitag, Fernando Pereira, MEMM Tutorial

Fri, 10/1/2004: No reading group

Fri, 10/8/2004: Conditional random fields

Papers to read:
John Lafferty, Andrew McCallum, Fernando Pereira, "Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data," ICML 2001.