Python机器学习：评价分类结果008多分类问题中的混淆矩阵

多分类问题中的混淆矩阵

#十分类问题
import matplotlib.pyplot as plt
import numpy as np

数据

from sklearn import datasets
digits = datasets.load_digits()
X = digits.data
y = digits.target

分割数据集

from sklearn.model_selection import train_test_split
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.8)

使用OvR方式解决分类问题

#默认使用OvR方式解决分类问题
from sklearn.linear_model import LogisticRegression

log_reg = LogisticRegression()
log_reg.fit(X_train,y_train)
log_reg.score(X_test,y_test)

0.93115438108484

y_predict = log_reg.predict(X_test)

默认求解二分类问题，如果需要多分类问题求解则需要调整参数average='micro‘

#precision_score默认求解二分类问题，如果需要多分类问题求解则需要调整参数
from sklearn.metrics import precision_score
precision_score(y_test,y_predict,average='micro')
#recall 和 f1_score也有类似参数

0.93115438108484

混淆矩阵

from sklearn.metrics import confusion_matrix
confusion_matrix(y_test,y_predict)

array([[139,   0,   0,   0,   1,   0,   2,   0,   0,   0],
       [  0, 138,   0,   3,   0,   0,   3,   1,   2,   7],
       [  0,   2, 134,   6,   0,   0,   1,   0,   5,   0],
       [  0,   2,   2, 128,   0,   2,   0,   4,   1,   2],
       [  0,   1,   0,   0, 130,   0,   0,   8,   2,   1],
       [  0,   2,   0,   0,   1, 138,   1,   1,   1,   4],
       [  0,   1,   0,   0,   0,   0, 140,   0,   0,   0],
       [  0,   0,   0,   0,   0,   0,   0, 148,   2,   0],
       [  0,   5,   1,   0,   0,   3,   1,   0, 117,   3],
       [  0,   3,   0,   1,   0,   3,   0,   7,   1, 127]], dtype=int64)

绘制一下混淆矩阵

#绘制一下混淆矩阵
cfm = confusion_matrix(y_test,y_predict)
plt.matshow(cfm,cmap=plt.cm.gray)

继续处理，我们关注的是误判的

#继续处理
row_sums = np.sum(cfm,axis=1)#行方向求和
erro_matrix = cfm / row_sums # 计算错误比例
#把对角线数据（预测正确的设置为0）
np.fill_diagonal(erro_matrix,0)
erro_matrix

array([[0.        , 0.        , 0.        , 0.        , 0.00704225,
        0.        , 0.0141844 , 0.        , 0.        , 0.        ],
       [0.        , 0.        , 0.        , 0.0212766 , 0.        ,
        0.        , 0.0212766 , 0.00666667, 0.01538462, 0.04929577],
       [0.        , 0.01298701, 0.        , 0.04255319, 0.        ,
        0.        , 0.0070922 , 0.        , 0.03846154, 0.        ],
       [0.        , 0.01298701, 0.01351351, 0.        , 0.        ,
        0.01351351, 0.        , 0.02666667, 0.00769231, 0.01408451],
       [0.        , 0.00649351, 0.        , 0.        , 0.        ,
        0.        , 0.        , 0.05333333, 0.01538462, 0.00704225],
       [0.        , 0.01298701, 0.        , 0.        , 0.00704225,
        0.        , 0.0070922 , 0.00666667, 0.00769231, 0.02816901],
       [0.        , 0.00649351, 0.        , 0.        , 0.        ,
        0.        , 0.        , 0.        , 0.        , 0.        ],
       [0.        , 0.        , 0.        , 0.        , 0.        ,
        0.        , 0.        , 0.        , 0.01538462, 0.        ],
       [0.        , 0.03246753, 0.00675676, 0.        , 0.        ,
        0.02027027, 0.0070922 , 0.        , 0.        , 0.02112676],
       [0.        , 0.01948052, 0.        , 0.0070922 , 0.        ,
        0.02027027, 0.        , 0.04666667, 0.00769231, 0.        ]])

越亮犯错越多
把真值1预测成了9…

#越亮犯错越多
#把真值1预测成了9..
plt.matshow(erro_matrix,cmap=plt.cm.gray)
#之后就可以微调分类器。。