笛卡尔乘积下的FCM聚类研究

针对于模糊c-均值（FCM）算法在初始聚类中心选取不佳的情况下容易产生聚类错误划分的情况，从FCM算法出发提出了一种基于笛卡尔乘积的FCM聚类算法（C-FCM），并分析了加权指数m对聚类分析的影响。C-FCM将聚类提高到更高维的空间，有效地避免了FCM 对初值敏感及容易陷入局部极小的缺陷。客运专线列控（TCC）评估测试项目对C-FCM的检验结果表明，与传统FCM算法相比，C-FCM算法更准确，效果更佳，对解决邻站数据包的划分问题是可行、有效的。

FCM clustering algorithm based on Cartesian product

As to the inaccuracy partition caused by the improper choices of initial cluster centers when adopting the Fuzzy C-Means（FCM） clustering algorithm,in this paper,FCM clustering algorithm based on Cartesian product and the effect of weighting exponent m on the performance of fuzzy clustering are studied in FCM algorithm.C-FCM improves the clustering to a higher dimension and avoids the disadvantages of local optimality and initialization dependence.The testing conclusion in the TCC testing project which used C-FCM shows that, the proposed method is more accurate and efficient than FCM and it is feasible and effective for the classification of packets from adjacent station.