Intensive study of a major sub-field such as neural networks, expert systems, machine learning/tutoring, natural language processing, pattern recognition, robotics, or others.
Course Type: ELECTIVE
SPECIFIC COURSE GOALS
TBD
LIST OF TOPICS COVERED
Introduction
Definitions
AI, Expert System, Rule-Based Expert System (RBES)
How an RBES works
Brief history of RBES
Applications of RBES
Foundation of REBES: Rule-Based Production Systems (RBPS)
Production system formalisms
Operational principles of RBPS
Evaluation of RBPS
Advantages
Disadvantages
Inference Engines (Automated RBPS)
Search
Chaining
Conflict resolution
Success and failure
Development of RBES using CLIPS (NASA’s RBES tool)
Tutorial on CLIPS
Preconditions
Stages
Problem selection
Knowledge acquisition: elicitation and induction
Knowledge representation: facts and rules
Design of the human interface
Design of the production system
Design of the explanation system
Iterative prototyping
Verification: consistency and completeness
Validation
Application
Problems and pitfalls
Fuzzy Logic
Representation of uncertainty
Abstraction as a solution
Bayesian logic as a solution
Certainty factors as a solution
Fuzzy logic as a solution
Tutorials on fuzzy logic
Classical Set Theory (Cantor)
Multi-Valued Logic (Lukasiewics)
Relationships: complement, containment, intersection, union
Formal definitions
Membership graphs: S, Z, and Pi
Linguistic Variables, Values, and Modifiers (Hedges)
Development of RBES Using Fuzzy CLIPS
Tutorial on Fuzzy CLIPS (an extension of CLIPS)
Design considerations
Preconditions for a “Fuzzy” solution
Methods of representing uncertainty in Fuzzy CLIPS
Major application areas for fuzzy expert systems
Advantages of Fuzzy Inference Control
Case Studies of Successfully Deployed Expert Systems
MACSYMA
MYCIN
XCON
PROSPECTOR
Evaluation of Expert Systems
Ethical issues in expert systems
Benefits of expert systems compared to human experts
