中文摘要：

为克服传统FCM只能对各类规模相同的样本聚类的不足，提出合影响力因子的FCMef聚类算法。对算法中的隶属度更新规则、目标函数、聚类中心更换规则进行证明。通过多组实验表明，FCMef聚类的效果比传统FCM要好，有较高的收敛精度，其收敛速度也明显大于FCM算法，当样本规模对比度较大时，表现尤为明显。

英文摘要：

The classic fuzzy c-mean（FCM） can＇t cope with situation of different sizes of samples. The method of attaching an effectiveness factor （HCMef） is proposed, and the update rules of the membership degree, the center of the cluster and the objective function are theoretically proved. The results of some experiments show that the clustering effect of FCMef is better than that of FCM because of its the higher convergence precision and the faster convergence speed. The larger scale samples are experimented, the better effect of FCMef cluster is obvious.