ML之LoR：基于LoR(逻辑回归)算法对乳腺癌肿瘤(9+1)进行二分类预测(良/恶性)

ML之LoR：基于LoR(逻辑回归)算法对乳腺癌肿瘤(9+1)进行二分类预测(良/恶性)

---

输出结果

Testing accuracy (10 training samples): 0.8685714285714285
Testing accuracy (all training samples): 0.9371428571428572

设计思路

数据集

可知该原始数据共有699条样本，每条样本有11列不同的数值：1列用于检索的id，9列与肿瘤相关的医学特征，以及一列表征肿瘤类型的数值。所有9列用于表示肿瘤医学特质的数值均被量化为1~10之间的数字，而肿瘤的类型也借由数字2和数字4分别指代良性与恶性。不过，这份数据也声明其中包含16个缺失值，并且用“?”标出。本代码中，对于存在缺失值的数据，都暂时子以忽略。

Number of Instances：699 (as of 15 July 1992)
Number of Attributes：10 plus the class attribute
Attribute Information：(class attribute has been moved to last column

#Attribute                             Domain                               
1 .Sample code number         id number
2 .Clump Thickness                  1一10
3 .Uniformity of Cell                  1一10
4 .Uniformity of Cell                  1一10
5 .Marginal Adhesion                1一10
6 .Single Epithelial Cell Size     1一10
7 .Bare Nuclei                           1一10
8 .Bland Chromatin                   1一10
9. Normal Nucleoli                    1一10
10 .Mitoses                               1一10
11 .Class:               (2 for benign,  4 for malignant)

核心代码

import numpy as np
intercept = np.random.random([1])
coef = np.random.random([2])

lx = np.arange(0, 12)
ly = (-intercept - lx * coef[0]) / coef[1]

from sklearn.linear_model import LogisticRegression
lr = LogisticRegression()
lr.fit(df_train[['Clump Thickness', 'Cell Size']][:10], df_train['Type'][:10])
print('Testing accuracy (10 training samples):', lr.score(df_test[['Clump Thickness', 'Cell Size']], df_test['Type']))

intercept = lr.intercept_
coef = lr.coef_[0, :]
ly = (-intercept - lx * coef[0]) / coef[1]

lr = LogisticRegression()
lr.fit(df_train[['Clump Thickness', 'Cell Size']], df_train['Type'])
print('Testing accuracy (all training samples):', lr.score(df_test[['Clump Thickness', 'Cell Size']], df_test['Type']))

intercept = lr.intercept_
coef = lr.coef_[0, :]
ly = (-intercept - lx * coef[0]) / coef[1]