采用K-means聚类，实现多维矩阵的聚类，并进行可视化展示（matlab）.rar

clc; clear; load('football_test.mat'); data=football_km; k=3;%设置聚类数目 max_iters=100; %设置迭代次数； initial_centroids = kMeansInitCentroids(data, k);%聚类中心初始化 [centroids, idx] = runkMeans(data, initial_centroids, max_iters);%运行k均值，返回聚类中心和类别 select=[1,36,64];%选择要展示的特征 biaoji=['bo';'r*';'k+';'g^'];%画图的标记 figure; %这是画图的部分 hold on; grid on; for i=1:k plot3(data(idx==i,select(1)),data(idx==i,select(2)),data(idx==i,select(3)),biaoji(i,:)); end [idx2,centroids2]=kmeans(data,k);