Knowledge-Systems Engineering
To be Offered Spring 2014
Technical Elective for ECE
Catalog Data: 

Graduate Course Information


ECE 566 - Knowledge-System Engineering

Credits: 3.00

Course Website: D2L

UA Catalog Description:

Course Assessment:

Homework:  3 assignments

Project:  1 class project

Activities: A few class activities

Exams:  1 Midterm Exam, 1 Final Exam

Grading Policy:

Typically: 25% Midterms,

                  25% Final Exam,

                  20% Homework,

                  10% Activities,

                  20% Class Project. 

Course Summary:

Knowledge-systems are intelligent systems that totally or partially involve computational representation and processing of knowledge. Objectives of this class are to:

(i)         Introduce students to the principles and techniques for engineering and developing knowledge systems.

(ii)        Teach the alternative computational structures and methods for representation of knowledge,

(iii)       Teach procedures and algorithms for computational processing (of knowledge structures) including automated reasoning and inference from knowledge, learning new knowledge, handling uncertain information, and complex knowledge-based decision making,

(iv)      Discuss alternative system architectures (engines) for knowledge-based systems.  

Graduate Standing

“Artificial Intelligence, A Modern Approach”, Latest Edition by Stuart Russell and Peter Norvig, Prentice-Hall.

Additional reading as provided/assigned by Instructor

Course Topics: 


1. Knowledge Representation in first order logic

·         First Order Logic

·         Matching and Unification

2. Expert Systems

·         Rule Firing, forward and backward reasoning

3.  Automated Planning and Problem Solving

·         Total Order problem solvers

·         Least Commitment Planning

·         Hierarchical Problem solving


4. Search Methods

·         Depth-first Search, Breadth-first Search 

·         Heuristic Search, A* algorithms

·         Hill Climbing

5. Structured Knowledge Representation

·         Representing Knowledge using Frames, and Semantic Networks

·         First Order Logic Correspondence

·         Matching

·         Inheritance, Defaults, and Automated Inference

6. Constraints, Constraint Networks, Constraint Satisfaction

·         Node and arc Consistency, Compound Labeling

·         Constraint Satisfaction algorithms

·         Problem Reduction, Look ahead, Back Jumping

·         Interval Constraints, Interval calculus

·         Algorithms for Interval Constraint Satisfaction

7. Knowledge sharing and collaborative problem solving

·         Blackboard model

·         Distributed agents, Peer to peer systems

·         Semantic commitments, Ontologies, and knowledge translation

8. Genetic Algorithms

·         Genetic representation of knowledge

·         Genetic operators, Fitness functions

·         Genetic search and optimization, Genetic learning

9. Bayesian Probabilistic Networks

·         Fundamentals from Probability Theory

·         Likelihood Vectors, and Conditional Probability Matrices

·         Hierarchical Propagation of Evidence

·         Computational Algorithms for General Belief Networks

10. Dempster-Shafer Theory of Evidence

·         Belief Interval Representations for uncertainty

·         Evidence Accumulation and Propagation

11. Knowledge-based Decision systems

·         Intro to Utility Theory, Utility functions

·         Decision networks

·         Decision-theoretic knowledge systems

·         Sequential decision problems, value iteration

Class/Laboratory Schedule: 

Lectures:  Multi-media streamed via the Web

6 classroom sessions, 75-minutes each

Assignments, Projects and Exams delivered via the course web site

Prepared by: 
Michael Marefat
Prepared Date: 
April 2013

University of Arizona College of Engineering