Sisteme inteligente de control si clasificare
FISA DISCIPLINEI

Anul universitar 2011- 2012


   Departament Home

Cod: MMIA221
Titular curs: prof. univ. dr. I. Iancu
Forma de invatamant: Master
Ciclul: Anul II, Semestrul 2
Curs: 28h, Laborator: 28h
Nr. credite: 6
Profil: informatica
Specializare: informatica
Tip disciplina: obligatorie
Categoria formativa: de specialitate
Obiective:
  • Cunoasterea principalelor sisteme de control inteligent bazat pe rationamentul fuzzy
  • Familiarizarea cu cateva sisteme inteligente, realizate prin diverse combinatii intre sistemele fuzzy, cele neuronale si cele evolutive
Continutul cursului:

      I. Sisteme fuzzy
    1. Multimi fuzzy
      1. Notiuni de baza
      2. Operatii: fundamentale, bazate pe t-operatori
      3. Numere fuzzy
      4. Relatii fuzzy
    2. Metode de rationament
      1. Implicatii fuzzy
      2. Variabile lingvistice
      3. Rationament aproximativ
    3. Sisteme de control logic fuzzy
      1. Fuzzificare si defuzzificare
      2. Tipuri de sisteme: Mamdani, Tsukamoto, Sugeno, Larsen
      II. Sisteme hibride
    1. Sisteme neuro-fuzzy
      1. Inferenta fuzzy prin retele neuronale hibride
      2. Implementarea regulilor fuzzy prin retele neuronale
      3. Ajustarea parametrilor de control fuzzy
      4. Modelul ANFIS
      5. Modelul MATCH-DFZ
    2. Retele neuronale evolutive
      1. Evolutia parametrilor
      2. Evolutia arhitecturii
      3. Invatare evolutiva
      III. Sisteme de clasificare fuzzy
    1. Clasificare prin metode evolutive
      1. Codificarea si generarea regulilor initiale
      2. Determinarea clasei consecinta si a certitudinii asociate
      3. Clasificarea datelor de antrenament
      4. Evolutia populatiei de reguli
    2. Clasificare prin metode neuro-evolutive
      1. Structura retelei neuronale
      2. Determinarea evolutiva a regulilor de clasificare
      3. Rafinarea regulilor extrase

Discipline anterioare cerute:
  • Algoritmi genetici
    Cod: I3507
  • Inteligenta artificiala
    Cod: I2303
  • Metode de calcul neuronal si evolutiv
    Cod: MMIA211
Discipline anterioare recomandate:
  • Medii de programare in Inteligenta Artificiala
    Cod: I3603
Forma de evaluare: examen
Continutul laboratorului:
  1. Implementarea cate unui sistem intelligent din cele trei tipuri: control fuzzy, sisteme hibride, clasificare fuzzy
Bibliografie:
  1. Iancu I.- Retele neuronale, Editura “Universitaria”, Craiova, 2007
  2. Iancu I. - Algoritmi genetici, Editura SITECH, Craiova 2008
  3. Iancu I. - Calcul evolutiv, Editura Universitaria, Craiova 2009
  4. Iancu I. - Incertitudine si imprecizie in sisteme inteligente, editura Universitaria, Craiova, 2002
  5. Ebrahim Elafi A., Haque R., Esmel Elalami M. – Extracting rules from trained neural network using GA for management E-bussines, Applied Soft Computing 4(2004), 65-77
  6. Fuller R.- Neural-Fuzzy Systems, Abo Akademi University, 1995
  7. Lee K.-M., Kwak D.-H., Leekwang H. – Tuning of Models by fuzzy neural network, Fuzzy Sets and Systems 76 (1995), 47-61
  8. Lee K.-M., Kwak D.-H., Leekwang H.- Techniques in Fuzzy Inference Neural Network for Fuzzy Model Improvement and their Application, Fuzzy Theory Systems . Techniques and Applications, vol. 3, Academic Press, 1999, 1241-1263
  9. Lee K.-M.., Leekwang H.- Fuzzy Inference Neural Network for Fuzzy Model Tuning, IEEE Transactions on Systems, Man and Cybernetics, vol 26, no. 4, August 1996, 637-645
  10. Nakashima T. – Fuzzy Genetics_Based Machine Learning for Pattern Classification, Thesis, Osaka Prefecture University, Osaka, Japan, 2000 /li>
  11. Pena-Reyes C. A., Sipper M. – A fuzzy- genetic approach to breast cancer diagnosis, Artificial Intelligence in Medicine, 17(1999), 131-155
  12. Rocha M., Cortez P., Nevez J. – Simultaneous Evolution of Neural Network Topologies and Weight for Classification and Regression, in: J. Cabestany, A. Prieto and D. V. Sandoval, IWANN 2005, LNCS 3512, pp. 59-66
  13. Yao Xin – Evolutionary Artificial Neural Network, in A. Kent (Ed.): Encyclopedia of Computer Science and Technology, Vol. 33, pp. 137-170, Marcel Dekker Inc., New York, 1995
  14. Yao Xin – Evolving Artificial Neural Network, Proc. of IEEE , vol 87, no. 9, 1999, pp.1423-1447
  15. Yao Xin, Liu Yong- A New Evolutionary Systems for Evolving Artificial Network, IEEE Transactions of Neural Networks, vol. 8, no. 3, may 1997, pp. 694-713
  16. Zadeh L. A. – Fuzzy Logic, Neural Networks and Soft Computing, Communications of the ACM, vol. 37, no. 3, 1994, pp. 77-84
Ultima actualizare: octombrie 2011