Cybernetics and Artificial Intelligence

The course is a substitute for:

Cybernetics and Artificial Intelligence (XE33KUI)

Lecturer:

Tutor:

Supervisor:

Department of Cybernetics

Synopsis:

A very general course enabling to understand the goals and principles of cybernetics and artificial intelligence and to organize different topics studied in the branch of study within a unified framework. Main principles of information theory, control engineering, decision making and knowledge engineering are overviewed in a reasonable extent.

The unifying conceptual approach to many diverse parts of cybernetics and artificial intelligence is the key feature of this course.

Requirements:

Syllabus of lectures:

1. Systems and models in cybernetics. Cybernetics, system theory and artificial intelligence

2. Control theory as a part of cybernetics. Feedback control in cybernetics

3. Basics in information theory: basic terminology, signal, coding, information

4. Joint and conditional entropy and its properties. Mean joint entropy

5. Communication channel and its capacity. Codes and coding. Principle of maximum entropy

6. Decision making under uncertainty and risk Statistical and Bayesian decision making

7. Theory of games. Minimax principle

8. Pattern recognition and perception. Feature-based and syntactic classifiers. Cluster analysis

9. Goals of artificial intelligence. Task representation, state space search

10. Logic from the point of view of artificial intelligence. Formalization of problem solving by logic

11. Knowledge representation. Formal languages, automata, Turing machines

12. Heuristic knowledge. Expert systems and control. Distributed knowledge-based systems

13. Knowledge engineering and knowledge acquisition. Adaptive and learning systems

14. Applications of artificial intelligence: robotics, problems of system integration, diagnostics

Syllabus of tutorials:

1. Introduction, safety rules. Turing's test. Formulation of the individual tasks to be solved

2. Demonstration of AI systems in the Gerstner Lab

3. Demonstration of AI systems in the Center of Machine Perception

4. Information theory. Information measures. Signal, coding, examples

5. Entropy and its application. Estimates of entropy. Principle of maximum entropy

6. Capacity of a discrete and continuous channel. Information and thermodynamic entropy

7. Independent and correlated events. Statistical characteristics and their application in decision making. Bayesian decision making

8. Solving individual tasks I

9. Solving individual tasks II

10. Presentation of results

11. Logic as a tool for representing knowledge. What can be done by using Prolog?

12. Software tools of machine learning, neural nets and genetic algorithms

13. Expert systems I

14. Expert systems II

Study Objective:

Study materials:

[1] Rich, E., Knight, K.: Artificial Intelligence. Mc-Graw Hill, 1991

Note:

Further information:

No time-table has been prepared for this course

The course is a part of the following study plans: