MA C137 Artificial Intelligence in Game Design

CATALOG COURSE DESCRIPTION

This course explores the concepts and techniques of the use of artificial intelligence in electronic game design and production. Topics include: the use of neural nets and genetic algorithms; giving the appearance of intelligence by using “smart” search, pursuit and avoidance algorithms as well as by “cheating” or providing the opponent characters with more information than they could realistically be aware of; Turing Tests; and the techniques of modeling a variety of behavioral styles and levels of aggression.

COURSE OBJECTIVES

Upon completing this course, the student will be able to:

1. Appraise the algorithms most commonly used in the industry for the simulation of
   “smart” search, pursuit and avoidance.
2. Relate the history and future projections of artificial intelligence in game programming
   and design.
3. Construct a reasonably intelligent opponent, in C++, using a variety of techniques.
4. Compare the use of neural nets, genetic algorithms, rule-based, intelligent heuristics,
   and “cheating” techniques in the development of smart opponents.
5. Appraise the techniques and algorithms most commonly used in the industry for the
   simulation of aggression, behavioral styles, and unpredictability.
6. Construct a simple, but intelligent opponent in C++ that can learn from its mistakes,
   using a variety of techniques.
7. Evaluate the use of artificial intelligence in existing games.

DETAILED TOPICAL OUTLINE

1. Introduction
   1. History of Artificial Intelligence in Game Programming and Design
   2. Current Trends and Future Projections
      
2. The Simulation of Behavioral Styles
   1. Algorithms for Aggression and Unpredictability
   2. Algorithms for Search, Pursuit and Avoidance
   3. Machine Learning
      
3. Artificial Intelligence in the Design of Smart Opponents
   1. Rule-based Approaches, Intelligent Heuristics, and Expert Systems
   2. Genetic Algorithms
   3. Neural Nets
   4. Turing Tests
      
4. Application
   1. Using C++ and Available Objects to Simulate Intelligence
   2. Criteria for the Evaluation of Artificial Intelligence in Existing Games

ASSIGNMENTS AND METHODS OF EVALUATION

1. Reading Assignments: The textbook reading assignments will be discussed in
   class in order to assist the student in developing a sequential theoretical
   understanding of the techniques and processes required to accomplish the
   computer based laboratory assignments.
   
2. Typical Writing Assignments: Write a critique of the use of artificial
   intelligence in an existing commercial game design.
   
3. Typical Outside Assignments: Research current trends in the industry, or visit
   a production studio and observe the processes of the use of artificial
   intelligence in game design.

REQUIRED TEXTS

Tricks of the Windows Game Programming Gurus by Andre Lamothe, MacMillian Publishing Company.

TYPICAL EVALUATION AND GRADING SCALE

Students will be evaluated based on critiques, game design projects, written quizzes, and a written exam. In order to achieve a grade of "C" in this course, the student must complete all assigned design projects.

Typical evaluation criteria may include:

Game AI Critiques 10%
Game AI Projects 50%
Quizzes 30%
Final Exam. 10%
Total 100 %

A typical grading scale is:

90-100% A
80-89% B
70-79% C
60-69% D
0-59% F

10.25.2000

  Courses Index