Loading...

Fuzzy wavelet modeling using data clustering

Sadati, N ; Sharif University of Technology | 2007

349 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/CIDM.2007.368861
  3. Publisher: 2007
  4. Abstract:
  5. In this paper, a novel approach for tuning the parameters of fuzzy wavelet systems which are used for modeling of nonlinear and complex systems is proposed. In fuzzy inference system, each fuzzy rule is analogous to a wavelet basis function multiplied by a coefficient. Using clustering techniques, the center of these basis functions are located in the detected center of clusters. In this way, not only the approximation accuracy is increased, but also the number of unknown parameters is decreased. The feasibility of the proposed method is shown by modeling two highly nonlinear functions. The comparison of the results using the proposed approach, with the previous schemes, shows the effectiveness and superiority of this algorithm. © 2007 IEEE
  6. Keywords:
  7. Approximation theory ; Fuzzy inference ; Fuzzy rules ; Large scale systems ; Mathematical models ; Wavelet transforms ; Data clustering ; Fuzzy wavelet systems ; Nonlinear functions ; Wavelet basis function ; Data structures
  8. Source: 1st IEEE Symposium on Computational Intelligence and Data Mining, CIDM 2007, Honolulu, HI, 1 April 2007 through 5 April 2007 ; 2007 , Pages 114-119 ; 1424407052 (ISBN); 9781424407057 (ISBN)
  9. URL: https://ieeexplore.ieee.org/document/4221285