python 自编FCM模糊聚类

导入包

from sklearn import datasets
from sklearn.utils.linear_assignment_ import linear_assignment
import seaborn as sns
import matplotlib.pyplot as plt
import copy
from sklearn.metrics import confusion_matrix
from sklearn import metrics
import pandas as pd
import numpy as np

导入数据 iris

iris = datasets.load_iris()
data = iris.data
y = iris.target

自编FCM

def Dist(data1, data2):
    return np.linalg.norm(data1-data2, axis=0)  # 求行向量的欧式距离

def FCM(data, k, m, epsilon, dist):
    U0 = np.random.rand(data.shape[0], k)
    Sum = np.sum(U0, axis=1)  # 求每行的和
    Lins2 = Sum.reshape(len(data), 1)
    U = np.divide(U0, Lins2)  # 保证每行之和为1
    Error = 5
    while Error > epsilon:
        U_m = np.power(U, m)
        C = np.transpose(np.dot(data.T, U_m)/U_m.sum(axis=0))  # 更新类中心