Prerequisite: COSC 310
An introduction to the field of artificial intelligence, i.e., the study of ideas that enable computers to process data in a more intelligent way than conventional practice allows. Covers many information representation and information processing techniques. Explores the underlying theory including matching, goal reduction, constraint exploration, search, control, problem solving, and logic.
Course Outcomes
Upon successful completion of this course, the student should be able to
- Discuss AI terminology, progress, and issues
- Assess strength and weakness of blind search algorithms.
- Assess strength and weakness of heuristic search algorithms.
- Solve problems regarding knowledge and reasoning.
- Formulate and solve simple problems in logic inference.
- Assess strength and weakness in game searching algorithms.
- Assess strength and weakness in AI algorithms in planning.
- Formulate and solve problems in machine learning.
- Formulate and solve problems in vision or natural language processing.
- Understand the basics of robotics or expert systems.
- Write programs in functional or logical paradigms.
- Implement software solutions to AI problems.
Detail Course Outline
A. Introduction — 3 hours
a. What is AI
b. The foundations of AI
c. The History of AI
d. Intelligent agents
e. Agent based system
B. Search — 6 hours
a. Searching for solution
b. Uninformed/Blind search
c. Informed/ Heuristic search
d. A* search
e. Hill-climbing search
f. Genetic algorithms
g. Constraint satisfaction problems
C. Game — 5 hours
a. Games
b. Optical decision in games
c. Minimax algorithm
d. Alpha-Beta pruning
e. Imperfect real time decision
f. Games that include an element of chance
D. Logic — 6 hours
a. Knowledge-based agents
b. Syntax of First Order Logic
c. Semantics of First Order Logic
d. Reasoning patterns in propositional logic
e. First order logic
f. Inference in first order logic
g. Unification and lifting
h. Forward and backward chaining
i. Resolution
E. Planning — 5 hours
a. The planning problem
b. Planning with state space search
c. Partial order search
d. Planning with proportional logic
e. Planning and acting in the real world
F. Learning — 5 hours
a. Learning from observation
b. Knowledge in learning
c. Statistical learning methods
d. Reinforcement learning
Choose any two of the following:
G. Robotics — 5 hours
a. Robotics hardware
b. Perception
c. Planning to move and moving
d. Software architecture
H. Vision — 5 hours
a. Digitization
b. Low-level processing
c. Noise removal
d. Feature detection
e. Segmentation and the Hough transformation
f. Recovering 3D information
g. The waltz algorithm
h. Active vision
i. Object recognition
j. Scene recognition
I. Natural Languages — 5 hours
a. Signal processing
b. Syntax
c. Parsing
d. Semantics
e. Meaning
f. Pragmatics
g. Natural language generation
J. Expert Systems — 5 hours
a. Examples
b. History
c. Advantages of expert system
d. AI as an experimental discipline
Midterm Exams — 2 hours
Total = 42 hours
Final Exam: During Final Exam week
Evaluation Methods
The final grade for the course is determined as follows:
Midterm exam: 15%
Final Exam: 25%
Homework: 10%
Projects: 30%
Quizzes: 20%
Grading Scale: 90-100% = A, 80-89% = B, 70-79% = C, 60-69% = D, 0-59% = F
Attendance:
The attendance policy will conform to the universitywide attendance criteria.
Required Textbook:
Stuart Russell and Peter Norvig. Artificial Intelligence: A Modern Approach. 2nd Edition. Prentice Hall, Jan 2003, ISBN 0130803022
