The Application of Clustering Optimization in Data Mining Based on Multiple Kernel Function FCM Algorithm

Mono-nuclear kernel function is presented in this paper based on the fuzzy c-means clustering algorithm for data clustering to do the improvement in the field of data mining, puts forward the fuzzy c-means clustering algorithm based on multiple kernel function (MKFCM) algorithm. Under fully unsupervised learning method, a set of Gaussian kernel function combination are assigned different weights resolution to a new multiple kernel function, through application of the single fixed kernel function of the fuzzy c-means clustering algorithm (KFCM algorithm) has obvious advantages and reliability in the clustering class compared. The simulation results indicate that the MKFCM algorithm can be size and a few kinds of density difference very big bunch of good classification, compared KFCM algorithm of the phenomenon of the overlap between in data clustering have significant performance advantages, also has a better application prospect in the field of data mining.