PSU CS410/510SPS: Scheduling, Planning, and Search

Date: Spring, 1999
Time: Tuesday, Thursday 1400-1550
Instructor: Bart Massey <bart@cs.pdx.edu>
Location: TBA

This is a tentative syllabus. Some changes may occur before and/or after the course begins.

Summary

This course explores methods for the solution of constraint satisfaction and related problems using search techniques, in the context of real-world problems such as resource-bounded scheduling, enterprise planning, classical planning, and one- and two-player games. The emphasis is on coding projects, and on reading and reporting on selected literature.

The course is, of necessity, an overview of the area: there is far more material here than can be thoroughly covered in the time available. A student project and presentation will provide an opportunity for focus on a particular topic of individual interest.

The course announcement may be helpful in understanding the intent of the course.

Objectives

• Be able to identify and formulate real-world search problems.
• Be able to use some standard methods for determining the computational difficulty of real-world problems.
• Understand and be able to apply standard search techniques in the solution of realistic problems.
• Be familiar with standard formulations of planning and scheduling problems, and understand the role of search in their solution.

Course Materials

Michael Garey and David Johnson
Computers and Intractability: A Guide to the Theory of NP-Completeness
W H Freeman & Co.
ISBN: 0716710455
$29.95

I won't lecture out of this more than once or twice, but it is indispensable as a reference, both during the course and afterward. The title is a bit intimidating, but take heart: this book is a practical nuts-and-bolts guide to separating tractable problems from intractable ones, including a large number of real-world examples.

Coursework

Each student will be expected to do all assigned reading, to present a paper in class, and to do one or more coding homework assignments plus a small final coding project.

This is a seminar-style course: I will attempt to substitute discussion for lecture to the extent that I find it feasible. It is important that students come to class prepared to discuss the material.

Prerequisites

• A fluency in some reasonable programming language (C preferred).
• Familiarity with algorithms and data structures at the level of PSU CS350 or equivalent.
• Undergraduates by instructor consent only, please. (Interested undergraduates are very much encouraged to send me e-mail.)

Schedule