模糊C均值（FCM）python实现

模糊C均值聚类算法是一种常用的无监督分类方法，广泛的应用于数据挖掘领域中，主要融合了K均值聚类算法和模糊理论，算法步骤如下所示：

Python实现

import torch
import random
import numpy as np
from sklearn import datasets


def FCM(data, N_way, iteration):
    # data : N*S, N为样本数，S为特征维数,tensor
    # N_way : 类别数
    #iteration:迭代次数
    #初始化
    N = data.shape[0]
    U = torch.rand(N_way, N)       #C*N,C为类别,N为样本
    for i in range(N):
        U[:,i] = U[:,i]/(U[:,i].sum())
    index = range(N)
    C = data[random.sample(index, N_way),:]        #C*S,C为类别，N为样本
    D = torch.zeros(N_way, N)                      #样本和类中心的距离矩阵


    count = 0
    #循环迭代
    while True:
        #计算聚类中心
        u = 1/U.sum(dim=1).reshape((N_way,1))
        c = torch.mm(U,data)
        C = c*u
        #更新划分矩阵
        for i in range(N_way):
            c_i = C[i,:]
            c_i = c_i.unsqueeze(0).expand(N, -1)
            d = (c_i - data)*(c_i - data)
            d = d.sum(dim=1)
            D[i,:] = d
        for i in range(N):
            mid = D[:,i]
            U[:,i] = D[:,i].sum()/D[:,i]
            U[:, i] = U[:, i] / (U[:, i].sum())
        #判断是否满足终止条件
        if(count>iteration):
            break
        count += 1
        #或者根据U（count+1）-U(count)判断
    return U,C


if __name__ == '__main__':
    iris = datasets.load_iris()
    n_sample, n_feature = iris.data.shape
    data = iris.data
    data = data.astype(np.float32)
    data = torch.from_numpy(data)
    U,C = FCM(data,3,100)
    label = iris.target
    count = 0
    for i in range(150):
        pre = torch.max(U[:,i],0)[1]
        print(pre)
        if pre==label[i]:
            count += 1

    print(count)