Course Syllabus

CS-480: Artificial Intelligence: Planning and Action

Course Description: Theory and practice of a variety of techniques for developing agents that can plan to achieve goals in open-ended environments. These include techniques of heuristic search, predicate logic with situation calculus, probabilistic reasoning, decision theory, and reinforcement learning, with application to software and robotic agents.

Textbook: Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach, Prentice Hall Publishers, 2nd Edition, ©2003

Prerequisites: CS 331 and MATH 474 or equivalents.

Course Objectives: By the end of the course, the successful student will be able to:

• Understand the Turing test and its implications for the study of intelligence.
• Explain the concepts of optimal reasoning, human-like reasoning, optimal behavior, human-like behavior.
• Develop "PAGE" descriptions of an agents and determine which agent type is applicable.
• Formulate an efficient problem space for a problem expressed in English by expressing that problem space in terms of states, operators, an initial state, and a description of a goal state.
• Describe the problem of combinatorial explosion and its consequences.
• Select an appropriate brute-force search algorithm for a problem, implement it, and characterize its time and space complexities.
• Select an appropriate heuristic search algorithm for a problem and implement it by designing the necessary heuristic evaluation function.
• Describe under what conditions different heuristic algorithms guarantee optimal solution.
• Implement minimax search with alpha-beta pruning for some two-player game.
• Formulate a problem specified in English as a constraint-satisfaction problem and implement it using a backtracking algorithm.
• Use first-order logic to represent the knowledge in some domain (translated from an English description).
• Apply Bayes theorem to determine conditional probabilities.
• Formalize a planning problem.
• Implement a progressive, regressive, and partial-order planning systems.
• Explain the differences among three main styles of learning: supervised, reinforcement, and unsupervised.
• Implement algorithms for reinforcement learning.
• Understand motion planning methods for a robot with simple kinematics.
• Describe different robotic control architectures and understand the advantages of each.

Class Meetings: Meetings consist of lecture, discussion, problem solving, discussion of some homework solutions, and exams. Regular class attendance is essential and you are expected to do the reading and to be prepared to actively participate in class activities.

Computer Access: This course requires that you have ready access to the WWW so that you can download course materials and communicate via email with your instructor. If you are registered for this course, you should have access to this course's Blackboard site (see http://blackboard.iit.edu). All course documents and assignments will be posted on this site, as well as important announcements from time to time. Please confirm asap that you can access the site.

Assignments:

Assignments (7 total)	50%
Midterm Exam	20%
Final Exam	30%

Grade intervals: A [85,100]; B [70,85); C [65,70); D [50,65); E [0,50).
No late assignments will be accepted!

Ethics: Any behavior on the homework assignments or exam that is adjudged to be copying or other cheating will result in an immediate zero on the assignment for all parties involved, and possible failure in the class and notification of the Dean's Office.

Communication is critical to the success and satisfaction of the learning experience. Please take advantage of my office hours, e-mail and phone numbers to communicate any issues or questions with me.

Note: There will be no class on April 3, 5, and 10, due to the Passover holiday.

Approximate Course Schedule: