Gradient based fuzzy c-means (GBFCM) algorithm

Author

Park, Dong C. ; Dagher, Issam

Author_Institution

Intelligent Comput. Res. Lab., Florida Int. Univ., Miami, FL, USA

Volume

3

fYear

1994

fDate

27 Jun-2 Jul 1994

Firstpage

1626

Abstract

In this paper, a clustering algorithm based on the fuzzy c-means algorithm (FCM) and the gradient descent method is presented. In the FCM, the minimization process of the objective function is proceeded by solving two equations alternatively in an iterative fashion. Each iteration requires the use of all the data at once. In our proposed approach one datum at a time is presented to the network, and the minimization is proceeded using the gradient descent method. Compared to FCM, the experimental results show that our algorithm is very competitive in terms of speed and stability of convergence for large number of data

Keywords

convergence of numerical methods; fuzzy neural nets; fuzzy set theory; iterative methods; minimisation; pattern classification; self-organising feature maps; Kohonen network; clustering algorithm; convergence; fuzzy c-means algorithm; gradient descent method; iterative method; minimization; numerical stability; objective function; Clustering algorithms; Convergence; Equations; Iterative algorithms; Minimization methods; Neurons; Stability;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on

Conference_Location

Orlando, FL

Print_ISBN

0-7803-1901-X

Type

conf

DOI

10.1109/ICNN.1994.374399

Filename

374399