中文摘要：

在传统FCM算法基础上提出一种称为特征加权和优化划分的模糊C均值算法（WGIFP-FCM）.该算法结合了优化划分法和特征权重思想,得到特征权重向量迭代公式并在聚类处理过程中自适应地更新.仿真实验结果表明,该算法能科学地计算出样本特征权重向量,获得优于GIFP-FCM、WFCM和FCM的聚类效果,具有一定实际应用价值.

英文摘要：

This paper proposes a new algorithm based on the FCM, which is called feature weighting and optimal fuzzy partition of fuzzy c-means algorithm （WGIFP- FCM）. This algorithm combines the optimization partition method and the feature weight idea, obtains feature weight vector formula and updates it in the entire clustering process. Simulation experiments have proved that the proposed algorithm has gotten better clustering results than GIFP - FCM, WFCM and FCM while calculating the weighting vector scientifically, and so it will be available in practice.