基于乳腺癌数据集建立决策树分类器

• 发布日期：2019-10-22
• 难度：一般
• 类别：分类与预测、决策树
• 标签：Python、决策树、breast_cancer、乳腺癌数据集

1. 问题描述

基于乳腺癌数据集的决策树算法实例如下。

2. 程序实现

# 导入乳腺癌数据
from sklearn.datasets import load_breast_cancer
cancer=load_breast_cancer()
#划分为训练集和测试集
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(cancer.data, cancer.target, test_size=0.3, random_state=42)
print(X_train.shape)
print(X_test.shape)

(398, 30)
(171, 30)

#模型训练
from sklearn.tree import DecisionTreeClassifier
clf=DecisionTreeClassifier(random_state=3)
clf.fit(X_train,y_train)
y_pred=clf.predict(X_test)
#准确率评估
train_score=clf.score(X_train,y_train)
test_score=clf.score(X_test,y_test)
print("train_score:%s" % (train_score))
print("test_score:%s" % (test_score))

train_score:1.0
test_score:0.929824561404

#接下来，本节以决策树深度max_depth为例，将模型深度与模型效果之间的关系画出来，直观地看到参数值与模型准确度之间的关系。
#定义“深度”参数选择函数
def score(d):
    clf=DecisionTreeClassifier(max_depth=d)
    clf.fit(X_train,y_train)
    train_score=clf.score(X_train,y_train)
    test_score=clf.score(X_test,y_test)
    return(train_score,test_score)
depths=range(2,15)
scores=[score(d) for d in depths]
train_score=[s[0] for s in scores]
test_score=[s[1] for s in scores]
#画图
import matplotlib.pyplot as plt
plt.figure()
plt.grid()
plt.xlabel('max depth of decision tree')
plt.ylabel('score')
plt.plot(depths,train_score,'.g-',label='train score')
plt.plot(depths,test_score,'.r--',label='test score')
plt.legend()
plt.show()
# 随着模型深度的增加，绿色实线代表的训练集准确度越来越高，而红色虚线代表的测试集准确度经过了上升下降波折，其中在深度为3时取得了最高值。