Recommended Texts

Causality, Judea Pearl, Cambridge University Press 2000.

Artificial Intelligence, Russell and Norvig, Prentice Hall 2003.

Reinforcement Learning, Sutton and Bartho 1999.

Overview

The field of Machine Learning aims to find general patterns in available data, and to use this information to improve the performance of information systems through experience. Central research issues are what kinds of information to seek from data, how to represent general information, and what to infer from given data. The course surveys the most common approaches to designing learning systems, including decision trees, Bayes Networks, neural nets and genetic algorithms. In general, the course offers an introduction to a variety of topics rather than an in-depth treatment of a specialized area.

Objectives of the Course

At the end of the course, students should be able to

- clearly formulate a learning problem

- be able to apply several standard tools in a given learning problem

- apply some theoretical analysis to evaluate the strengths and weaknesses of a learning system design

Class Format

Typically, we will begin with a lecture-style introduction to a new topic. Then we will discuss examples and applications, especially the results of students experimenting with learning software. We will also go over exercises, possibly with students taking the lead in presenting a solution.

Marking

Final and/or Project: 50%

Assignments: 30%

Scribe: 20%

Notes:

1. Scribe: I ask each student to write lecture notes for one week. The notes will then be posted on the web as a common resource.

a. Since we have a lot of students, there may be more than one scribe for a given week. In that case I suggest that you divide up the task by a) lectures (Tuesday vs. Thursday) and/or b) subject matter (e.g., split the reading in half and cover one what’s associated with your part of the lecture). Please indicate on your submission which part you have done.

b. The write-up will be due the Tuesday following you week that you are covering.

c. I will also ask the scribes for the week to present solutions to the assignments.

2. There will be weekly assignments.I encourage you to work in groups, but you should write up your results on your own. If you work with others, put down their names.

3. A project is not required. For those who want to do the extra work, I would grade them on the higher of {(Final 15%, Project 35%, Ass 30%, Scribe 20%), (Final 35%, Project 15%, Ass 30%, Scribe 20%)}. A C in this course requires a grade of at least 50% on the final exam.

Final Grades

Because of small sample size irregularities, the exact cutoffs for converting term marks into letter grades may depend on the class distribution to some extent. The following conversion table shows a typical conversion method.

A+	96-100%
A	86-95%
A-	80-85%
B+	78-79%
B	73-77%
B-	70-72%
C+	68-69%
C	63-67%
C-	60-62%
D	50-59%
F	0-49%

Handing In Written Work

1. Staple your paper.

2. Write down your name. It’s a good idea to put down your E-mail address and telephone number as well in case I have questions about your work.

3. Write down the names of the people who worked with you on this assignment (if any).

4. You should hand in your assignment in class on the day that it is due.

5. All submitted work should be written in pen, not pencil.

6. No late assignments will be accepted.

Studying for this Class

1. Read the assigned readings - at least twice.

2. It’s best to keep up with the regular readings and assignments. Failing an assignment is a serious warning sign, even if it doesn’t hurt your grade that much immediately.

3. Problems that you should address immediately before they lead to more difficulties:

- can’t understand examples worked out in lectures

- can’t do exercises

- don’t know what symbols mean.

4. I recommend that you study in groups and do assignments together. Just down put on your assignment who you worked with.

Students With Special Needs

Students who require accommodations in this course due to a disability affecting mobility, vision, hearing, learning, or mental or physical health are advised to discuss their needs with The Centre for Students with Disabilities, 291-3112 (Phone) or www.sfu.ca/student-services/disabilities.html .

Plagiarism

Plagiarism is an extremely serious academic offense, and will not be tolerated in this course. SFU’s Code of Academic Policy (http://www.sfu.ca/policies/teaching/t10-02.htm) states:

“Plagiarism is a form of academic dishonesty in which an individual submits or presents the work of another person as his or her own. Scholarship quite properly rests upon examining and referring to the thoughts and writings of others. However, when excerpts are used in paragraphs or essays, the author must be acknowledged using an accepted format for the underlying discipline. Footnotes, endnotes, references and bibliographies must be complete…

Plagiarism exists when all or part of an essay is copied from an author, or composed by another person, and presented as original work. Plagiarism also exists when there is inadequate recognition given to the author for phrases, sentences, or ideas of the author incorporated into an essay.

A draft paper, proposal, thesis or other assignment may be subject to penalty for academic dishonesty provided the instructor/supervisor has informed the student(s) before the work is submitted…

Penalties imposed by the University for academic dishonesty may include but are not limited to one or more of the following: a warning, a verbal or written reprimand, reassessment of work, failure on a particular assignment, failure in a course, denial of admission or readmission to the University, deregistration, forfeiture of University awards or financial assistance, suspension or permanent suspension from the University or revocation of a degree.”