Research on improvement of fuzzy clustering analysis

The traditional fuzzy clustering analysis can not effectively resolve the problems of different importance and relevance interfering of the characteristic attributes, aiming at these deficiencies, an improved algorithm which combined with AHP and Mahalanobis distance algorithm is presented. The improved algorithm not only solved the problems of the traditional algorithm, but also reduced the adverse effects of the subjective factors to Clustering Analysis, and the results of the analysis are more objective and accurate.