Syllabus - Probabilistic Models in Informatics

The Graduate University for Advanced Studies (SOKENDAI), Department of Informatics

The focus of this course is probabilistic models, which play an important role for the modeling of real-world data in various fields of informatics. This course deals with foundations and applications of probability theory, and discusses basic ideas of probabilistic models, such as latent variable models and Bayesian Networks.

Course Plan

1	10/19	Introduction to Probability
2	10/26	Probability Theory and Bayes Theorem
3	11/02	Random Variable and Probability Distribution
4	11/09	Multivariate Probability Distribution
5	11/16	Maximum Likelihood
6	11/30	Mixture Models
7	12/07	Model Selection
8	12/14	Hidden Markov Model (1)
9	12/21	Hidden Markov Model (2)
10	01/11	Graphical Models (1)
11	01/18	Graphical Models (2)
12	01/25	Bayesian Network
13	02/01	Sampling and Simulation
	02/08	*Cancel*
14	02/15	Summary

Report 1

Pick up some probabilistic phenomena around your life (coin tossing, more complex phenomena, etc.)
Perform simulation for generating data.
Assume some probabilistic models and estimate parameters from data.
Compare simulation and parameter estimation.
Report your problem, how you did simulation, how you assumed the model, and how you compare simulation and model.
Submit your report in Word or PDF.
Deadline is January 18, 2006.

Paper Reading Assignment

1	01/11	Duy Le	An adaptive color-based particle filter, Image Vision Computing, 2003
2	01/18	Masoomeh Torabzadeh	On the Self-Similar Nature of Ethernet Traffic, IEEE/ACM Trans. Networking, 1994
3	01/25	Sano Masanori	An Integrated Baseball Digest System Using Maximum Entropy Method, ACM Multimedia 2002
4	02/01	Nguyen Phuoc Tat Dat	Tagging English Text with a Probabilistic Model, 1994
5	02/15	Nguyen Luu Thuy Ngan	Probabilistic Reasoning for Entity & Relation Recognition, COLING'02

Pick up one paper, which may be a standard reading in your research field.
Report the title of the paper to me *one week before your presentation*, and also send me the copy of the paper if possible.
Describe how the paper takes advantage of probabilistic models.
15-min presentation, and 15-min discussion.

Syllabus - Probabilistic Models in Informatics - Second Semester 2005

Course Plan

Report 1

Paper Reading Assignment