CSEP 546 - Machine Learning

University of Washington School of Computer Science & Engineering

CSEP 546 - Data Mining / Machine Learning - Autumn 2017
Mondays 6:30-9:20pm in Johnson Hall room 075

Instructor: Pedro Domingos (pedrod at cs)
Office hours: Mondays 5:30-6:20pm & by appointment, CSE 648

TA: Kenton Lee (kentonl at cs)
Office hours: Mondays 5:30-6:20pm at CSE 021 & by appointment
TA: Alon Milchgrub (alonmil at cs)
Office hours: Mondays 5:30-6:20pm at CSE 220 & by appointment

Schedule

Week	Dates	Topics & Lecture Notes	Readings
1	October 2	Introduction (ppt, pdf), inductive learning (ppt, pdf)	Mitchell, Ch. 1; Duda, Ch. 1; paper
2	October 9	Decision trees (ppt, pdf)	Mitchell, Ch. 3; Duda, Ch. 8
3	October 16	Rule induction (ppt, pdf)	Mitchell, Ch. 10; Duda, Ch. 8
4	October 23	Instance-based learning (ppt, pdf)	Mitchell, Ch. 8; Duda, Ch. 4
5	October 30	Bayesian learning (ppt, pdf)	Mitchell, Ch. 6; Duda, Ch. 2 & 3
6	November 6	Neural networks (ppt, pdf)	Mitchell, Ch. 4; Duda, Ch. 6
7	November 13	Model ensembles (ppt, pdf)	Duda, Ch. 9
8	November 20	Learning theory (ppt, pdf)	Mitchell, Ch. 7; Duda, Ch. 9; paper
9	November 27	Support vector machines (ppt, pdf)	Duda, Ch. 5
10	December 4	Clustering and dimensionality reduction (ppt, pdf)	Duda, Ch. 10

Textbooks

R. Duda, P. Hart & D. Stork, Pattern Classification (2nd ed.), Wiley, 2001 (required).
Tom Mitchell, Machine Learning, McGraw-Hill, 1997 (required).
Pedro Domingos, The Master Algorithm, Basic Books, 2015 (recommended).

Assignments

There will be four assignments handed out on weeks 2, 4, 6, and 8; they are due two weeks later. Each assignment is worth 25% of the final grade. Submit online here.

Assignment 1: Clickstream Mining with Decision Trees, Rule Induction. Due October 23, 2017 at 12:01am.
Assignment 2: Collaborative Filtering and Bayesian Networks. Due November 6, 2017 at 12:01am.
Assignment 3: Neural Nets & Ensemble Methods. Due November 19, 2017 at 11:59pm.
Assignment 4: Learning Theory and SVMs. Due December 3, 2017 at 11:59pm.

Course Administration and Policies

Discuss course material on the forum.
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.

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