FPF:UINA510 Expert Systems Applications - Informace o předmětu

UINA510 Expert Systems Applications

Rozsah

2/2/0. 6 kr. Ukončení: zk.

Vyučující

Ing. Jiří Blahuta, Ph.D. (přednášející)
Mgr. Anna Krajčírová (cvičící)
RNDr. Jiří Martinů, Ph.D. (cvičící)
Mgr. Daniel Valenta, Ph.D. (cvičící)

Garance

Ing. Jiří Blahuta, Ph.D.
Ústav informatiky – Filozoficko-přírodovědecká fakulta v Opavě

Rozvrh

Po 8:05–9:40 LEI

• Rozvrh seminárních/paralelních skupin:

UINA510/A: St 13:05–14:40 B3b, A. Krajčírová

Předpoklady

Absolvování předmětu Expertní systémy v bakalářském studijním programu.

Omezení zápisu do předmětu

Předmět je nabízen i studentům mimo mateřské obory.

Mateřské obory/plány

• Computer Science (program FPF, CompSci-np)

Cíle předmětu

The goal of the course is to acquaint students with actual trends in development and implementation of expert systems based on different types of vagueness and also with methods of softcomputing. Practical exercises acquaint students with used empty knowledge-based and/or expert systems with emphasis on integration of fuzzy approaches. Inseparable part of the exercises is also design of knowledge or expert basis of a fuzzy rule based expert system for a priori determined problem in given empty fuzzy rule based knowledge-based or expert system.

Výstupy z učení

The student will be able to create and implement an expert system based on various types of uncertainties and using soft computing methods. From a theoretical point of view and in practical exercises, to design and implement a knowledge/expert base fuzzy rule expert system for a given problem and also in a fuzzy rule-oriented knowledge/expert system.

Osnova

• Contens:
  
  • 1. Subjective knowledge and mental models, language modeling, conditional production rules, set of rules as language model, vagueness of the language model.
  • 2. Vagueness of conditional rules formalized by means of probability level, expert systems with entropy such as MYCIN, EMYCIN and PROSPECTOR.
  • 3. Probability-based rule-based expert systems - FEL-EXPERT.
  • 4. Formalization of vagueness of language terms by means of fuzzy sets, basics of fuzzy mathematics, approximation of language models of fuzzy functions, fuzzification and defuzzification.
  • 5. Fuzzy logic, language variable, interpretation of fuzzy logic functions, fuzzy relations, approximative derivation, fuzzy rule Modus Ponens.
  • 6. Fuzzy approximation of multidimensional non-linear systems, approximative derivation based on Takagi- Sugeno rules.
  • 7. Cognitive analysis of rule base of knowledges, consistency test, diversification of knowledge base, ternary graph.
  • 8. Artificial neural multi-layer networks, adaptation procedures of ANN, fuzzy neural networks, identification of the fuzzy model Takagi-Sugeno.
  • 9. Automatic methods of structural and parametric identification of rule-based fuzzy models, fuzzy clustering methods.
  • 10. Basics of fuzzy regulation, linear regulators in case of feedback regulation, fuzzy regulators synthesis, Sugenobased regulators.
  • 11. Methods of structural and parametric optimalization of rule-based fuzzy models, optimalization methods based on evolution.
  • 12. Real-time expert system, knowledge management, intelligent regulators, knowledge-based adaptation of structure and parameters of regulators.

Literatura

• CASTILLO, Enrique et. al. Expert systems and probabilistic network models. New York: Springer, 2012. ISBN 9781461274810.

Výukové metody

Lectures, discussions, exercises, case studies.

Metody hodnocení

* 75% attendance in exercises, active participation * written test in the extent of the given literature and the content of seminars 30 points * implementation of selected methods of softcomputing, success rate 30 points from programming and 10 points from documentation * 40 points exam 50% all parts

Vyučovací jazyk

Angličtina

Informace učitele

http://is.slu.cz
Výuka probíha v blocích, přednášky a cvičení.

• Statistika zápisu (nejnovější)
  
• Permalink: https://is.slu.cz/predmet/fpf/zima2025/UINA510