Machine Intelligence
Winter 2008

Prerequisite: CS 165A

Credits: 4

Catalog course description:

The course covers the most important techniques of machine learning and includes discussions of: well-posed learning problems; artificial neural networks; concept learning and general to specific ordering; decision tree learning; genetic algorithms; learning sets of rules; Bayesian learning; analytical learning; and combining inductive and analytical learning.

Textbook:

T. M. Mitchell, Machine Learning, McGraw Hill, 1997

Reference:

S. Russell and P. Norvig, Artificial Intelligence: A Modern Approach, Prentice Hall, 1995

Course goals:

The course introduces fundamental concepts of machine intelligence (MI) in the context of machine learning (ML). In many respects, it is likely that ML will prove to be one of the core elements of successful MI. We will provide a thorough overview of ML and, in particular, we will use a framework of ML to introduce several core threads of MI, including search, knowledge representation and reasoning, and applications. A goal is to allow students to obtain a greater depth of understanding of some topic of interest.

Prerequisites by topic: AI programming, generalized search

Lectures: MW 2:00-3:15pm, Phelps Hall 3523

Discussions: M 4:00-4:50pm, Phelps 1444

Office hours: TTh 9:15-10:30am, , or by appointment

Contact information: (805) 893-2966, smithtr@cs.ucsb.edu

Teaching assistant: TBA

Course homepage: http://www.cs.ucsb.edu/~cs165b/

Requirements:

1. There will be a midterm exam and a final exam (close-book, close-notebook). The scope of the midterm exam includes topics covered in lectures and discussions before the exam; the final exam covers everything taught in the course with an emphasis on the materials studied after the midterm.
2. There will be a project and 6 homework assignments.
3. Copying (parts of) answers/programs in homework, project, or an exam will automatically result in a "Failure" grade for the course and a report to the department and the university.

Grading: Project: 20%; Midterm: 20%; Homework: 20%; Final: 40%

Deadlines and late penalties:

Each homework assignment will have a due date and there will be penalties for each additional date that the homework is late, as described on the homework pages.

Topics and relevant readings:

Topics	Chapters in the textbook	Hours (Approximate)
Course organization and introduction		1
Well-posed learning problems	Chapter 1	3
Concept learning and general to specific ordering	Chapter 2	3
Decision tree learning	Chapter 3	2
Artificial neural networks	Chapter 4	3
Bayesian learning	Chapter 6	3
Instance-based learning	Chapter 8	3
Genetic algorithms	Chapter 9	2
Learning sets of rules	Chapter 10	3
Analytical learning	Chapter 11	3
Combining inductive and analytical learning	Chapter 12	3