Function for Classification and Nonlinear Time-Series Prediction  

Oleh:

Kim, Min-Soeng ; Hong, Sung-Gi ; Lee, Ju-Jang

Jenis: Article from Article
Dalam koleksi: Final Program and Book of Abstracts: The 4th Asian Control Conference, September 25-27, 2002 (Sep. 2002), page 226-231.
Topik: Non Linear; TSK Fuzzy; Optimization; Iris Classification
Fulltext: AC021487.PDF (256.09KB)

Isi artikel

In this paper, a new evolutionary optimization scheme to design a TSK fuzzy model from data is proposed. The identification of the antecedent rule parameters is done via the evolutionary algorithm with the newly defined fitness function and the various evolutionary operators. The proposed fitness function considers not only mean squared error, but also calculates some penalty based on the overlapping of adjacent membership functions for the same input. The occurrence of the multiple overlapping membership functions, which is a typical feature of unconstrained optimization, is resolved with the help of the newly defined fitness function. The identification of the consequent parameters is made with the pseudo inverse method. Through simulations on iris classification and nonlinear time se- ries prediction problem, the proposed algorithm found a TSK fuzzy model which produces fewer errors than found in previous works.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali