Title :
Fuzzy topological map algorithms. A comprehensive comparison with Kohonen feature map and fuzzy C-mean algorithms
Author :
Lam, Yiu C. ; Cheung, Kwan F.
Author_Institution :
Dept. of Electr. & Electron. Eng., Hong Kong Univ., Hong Kong
Abstract :
The function of clustering algorithm is to obtain an optimal reduced set of vectors which can, in some optimal sense, represent a given set of data vectors. Different criteria are used in different algorithms. In this paper, a class of clustering algorithms known as the Fuzzy Topological Map (FTM) algorithm, which is generalized from Kohonen Feature Map (KFM) algorithm and Fuzzy c-Mean (FCM) algorithm, is presented. In particular, KFM algorithm captures the topological structure of the data set, whereas FCM algorithm considers the reduction of an objective function in some L2 sense. The FTM algorithm is formulated to subsume both criteria
Keywords :
data compression; fuzzy set theory; image coding; image recognition; FTM algorithm; Kohonen feature map; clustering algorithm; data set; fuzzy C-mean algorithms; fuzzy topological map algorithms; image compression; objective function; optimal reduced set; topological structure; Clustering algorithms; Convergence; Curve fitting; Inference algorithms; Minimization methods; Noise generators; Speech recognition; Springs; Topology; Vector quantization;
Conference_Titel :
Circuits and Systems, 1997. ISCAS '97., Proceedings of 1997 IEEE International Symposium on
Print_ISBN :
0-7803-3583-X
DOI :
10.1109/ISCAS.1997.608785
