CSE 515 - Statistical Methods in Computer Science

University of Washington Computer Science & Engineering

CSE 515 - Statistical Methods in Computer Science - Winter 2018

CSE Home

About Us

Contact Info

Instructor: Pedro Domingos
Office: Allen 648
Office hours: Tuesdays 11:00am-11:50am and by appointment TA: Yao Lu
Office: Allen 220
Office hours: Thursdays 11:00am-11:50am and by appointment

Class meets:
Tuesdays and Thursdays from 9:30 to 10:50 in EEB 037

Schedule

Week	Dates	Topics & Lecture Notes	Readings
1	January 4	Introduction, basics of probability and statistical estimation	Ch. 1, 2 & 17
2	January 9 & 11	Mixture models and the EM algorithm (EM notes)	Ch. 19
3	January 16 & 18	Hidden Markov models and Kalman filters	Ch. 6 (Sec. 6.1 & 6.2) & Ch. 15 (Sec. 15.4.1)
4	January 23 & 25	Bayesian networks and Markov networks	Ch. 3 & 4
5	January 30 & February 1	Variable elimination, junction trees and belief propagation	Ch. 9 - 11
6	February 6 & 8	Sampling-based inference	Ch. 12
7	February 13 & 15	Learning Bayesian networks	Ch. 16 - 18
8	February 20 & 22	Learning Markov networks	Ch. 20
9	February 27 & March 1	Dynamic Bayesian networks, particle filtering and relational models	Ch. 15 (Sec. 15.1 - 15.3) & Ch. 6 (Sec. 6.3 & 6.4)
10	March 6 & 8	Decision theory and Markov decision processes	Ch. 22 & 23

Textbook

D. Koller & N. Friedman, Probabilistic Graphical Models: Principles and Techniques, MIT Press (required).

Assignments

There will be four homework assignments, each worth 25% of the final grade. The homeworks include programming. Homework assignments will be handed out after weeks 2, 4, 6, 8, and will be due two weeks later.

Assignment 1, due at 12:01am on January 30th.
Data: vehicle.train, vehicle.test.
See also: Vehicle silhouettes dataset at the UCI Machine learning repository
Assignment 2, due at 12:01am on February 13th.
Data and code: Package.zip
Assignment 3, due at 12:01am on February 27th.
Data and code: seg.zip.
See Kschischang et al. for a detailed explanation of BP.
Assignment 4, due at 12:01am on March 13th.

Canvas

We will use UW Canvas to collect homework, keep the gradebook and make announcements.

Course Administration and Policies

Assignments will be done individually unless otherwise specified. You may discuss the subject matter with other students in the class, but all final answers must be your own work. You are expected to maintain the utmost level of academic integrity in the course.
Assignments may be handed in up to a week late, at a penalty of 10% of the maximum grade per day.
Discuss course material on the forum.

Computer Science & Engineering
University of Washington
Box 352350
Seattle, WA 98195-2350
(206) 543-1695 voice, (206) 543-2969 FAX