Prerequisites: A Data Structures course.
Recommended: One additional CPS course plus four semesters of mathematics.
Emphasis on the tools and techniques of the field. Major topics: PROLOG (an AI language); expert systems, including Bayesian nets; search techniques as used in problem solving and game playing; neural nets; natural language; planning; automated reasoning (brief, the emphasis is on the relevant elements of logic and the origins of Prolog).
Selected readings will be assigned from the Luger and Stubblefield textbook. This book is optional. The other books are required.
Luger and Stubblefield, Artificial Intelligence: Structures and Strategies for Complex Problem Solving, 3rd edition, Addison-Wesley, 1997.
Clocksin and Mellish, Programming in Prolog, 4th edition.
McClelland and Rumelhart, Explorations in Parallel Distributed Processing: A Handbook... (One chapter available by CoursePack later in the semester)
Midterm exam.
Final exam.
Project report.
Homework = 20%
Project = 30%
Midterm = 20%
Final Exam = 30%
Because there now exists an undergraduate Artificial Intelligence course at Duke, it is assumed that any undergraduate taking this course is prepared to take a full graduate course. For example, students may be held responsible for material learned on their own (but such extra material will be clearly indicated). The extra material has traditionally been two or three chapters in the Luger and Stubblefield book, for which there is inadequate time to cover in class. Several copies will be on reserve so that purchase of the book is a matter of convenience. I should remark that seniors and even juniors who are good students and who are prepared to work have successfully taken this course in the past.
There are class notes for each lecture that are made available in advance of the lecture.