中文说明：[L, C, D] = FKMEANS(X, k)，使用k均值算法将矢量矩阵x分割成k聚类。行x对应点，列对应变量。输出k×p矩阵包含聚类中心。K-means算法是很典型的基于距离的聚类算法，采用距离作为相似性的评价指标，即认为两个对象的距离越近，其相似度就越大。该算法认为簇是由距离靠近的对象组成的，因此把得到紧凑且独立的簇作为最终目标。k个初始类聚类中心点的选取对聚类结果具有较大的影响，因为在该算法第一步中是随机的选取任意k个对象作为初始聚类的中心，初始地代表一个簇。该算法在每次迭代中对数据集中剩余的每个对象，根据其与各个簇中心的距离将每个对象重新赋给最近的簇。当考察完所有数据对象后，一次迭代运算完成，新的聚类中心被计算出来。如果在一次迭代前后，J的值没有发生变化，说明算法已经收敛。
English Description：
[L, c, d] = FKMEANS (x, k), using the k-means algorithm for matrix vector x into k clusters. Rows correspond to x, columns correspond to variables. Output KXP matrix containing the cluster Center.Was typical of the k-means algorithm for clustering algorithm based on distance, using evaluation as a similarity index, which holds that the distance between two objects, the closer, the greater its similarity. The algorithm that cluster is composed of near objects, so separate clusters are compact and as the ultimate goal.K initial clustering centers and the selection has a greater impact on clustering results, as in the first step of the algorithm is randomly select any as the initial clustering centers of k object, to represent an initial cluster. In each iteration, the algorithm for each object remaining in the data set, by relating them to the various cluster-center of each object to be assigned to the nearest cluster. When examining all the data objects when you are finished