Python实现线性判别

Python实现线性判别，参考周志华机器学习第三章

 import numpy as npimport matplotlib.pyplot as pltmat=np.loadtxt('gua.txt')X1=mat[0:8,1:3]        #好瓜的数据X0=mat[8:,1:3]         #坏瓜的数据mean1=np.mean(X1, axis=0)   #计算每一列的平均值mean0=np.mean(X0,axis=0)mean1=mean1.reshape(2,1)mean0=mean0.reshape(2,1)cov0 = np.cov(X0,rowvar=False)        #坏瓜协方差矩阵cov1 = np.cov(X1,rowvar=False)        #好瓜协方差矩阵Sw=cov0+cov1w=np.dot(Sw**(-1),mean1-mean0)print(w)plt.scatter(X0[:,0],X0[:,1],label='0')plt.scatter(X1[:,0],X1[:,1],label='1')plt.plot([0,1],[0,-w[0]/w[1]],label='y')plt.show()