Abstract: Traditional FCM is sensitive with the initial cluster center and noise also easily converge to a local minimum values, which leads to low clustering accuracy. This article proposes a method that uses divide and conquer technique with equivalency and compatible relation concepts to improve the performance of the FCM clustering method. Experiment results demonstrate appropriate accuracy.

Key words: Fuzzy C-means Clustering（FCM）, divide and conquer, unsupervised clustering, micro-array data

摘要： 传统的快速聚类算法大多基于模糊C均值算法（Fuzzy C-means，FCM），而FCM对初始聚类中心敏感，对噪音数据敏感并且容易收敛到局部极小值，因而聚类准确率不高。建立使用分治策略解决聚类问题的算法架构，充分考虑数据本身特性并对传统的FCM算法进行改进，标准数据集的实验结果表明这种基于分治策略的FCM聚类算法较好地提高了算法的聚类准确率，加快了收敛速度。

关键词: 模糊C均值聚类, 分治策略, 无监督聚类, 微阵列数据