机器学习——决策树模型：Python实现

from sklearn.tree import DecisionTreeClassifier
X = [[1, 2], [3, 4], [5, 6], [7, 8], [9, 10]]
y = [1, 0, 0, 1, 1]
model = DecisionTreeClassifier(random_state=0)
model.fit(X, y)
print(model.predict([[5, 5]]))

from sklearn.tree import DecisionTreeRegressor
X = [[1, 2], [3, 4], [5, 6], [7, 8], [9, 10]]
y = [1, 2, 3, 4, 5]
model = DecisionTreeRegressor(max_depth=2, random_state=0)
model.fit(X, y)
print(model.predict([[9, 9]]))

# 1.读取数据与简单预处理
import pandas as pd
df = pd.read_excel('员工离职预测模型.xlsx')
df = df.replace({ 
'工资': { 
'低': 0, '中': 1, '高': 2}})
# 2.提取特征变量和目标变量
X = df.drop(columns='离职') 
y = df['离职']   
# 3.划分训练集和测试集
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=123)
# 4.模型训练及搭建
from sklearn.tree import DecisionTreeClassifier
model = DecisionTreeClassifier(max_depth=3, random_state=123)
model.fit(X_train, y_train) 

y_pred = model.predict(X_test)
print(y_pred[0:100])
# 通过构造DataFrame进行对比
a = pd.DataFrame()  # 创建一个空DataFrame 
a['预测值'] = list(y_pred)
a['实际值'] = list(y_test)
a.head()
# 如果要查看整体的预测准确度，可以采用如下代码：
from sklearn.metrics import accuracy_score
score = accuracy_score(y_pred, y_test)
print(score)
# 或者用模型自带的score函数查看预测准确度
model.score(X_test, y_test)

y_pred_proba = model.predict_proba(X_test)
print(y_pred_proba[0:5])
b = pd.DataFrame(y_pred_proba, columns=['不离职概率', '离职概率']) 
b.head()

y_pred_proba[:,1]

from sklearn.metrics import roc_curve
fpr, tpr, thres = roc_curve(y_test, y_pred_proba[:,1])

import matplotlib.pyplot as plt
plt.plot(fpr, tpr)
plt.show()

from sklearn.metrics import roc_auc_score
score = roc_auc_score(y_test, y_pred_proba[:,1])
print(score)

model.feature_importances_
# 通过DataFrame进行展示，并根据重要性进行倒序排列
features = X.columns  # 获取特征名称
importances = model.feature_importances_  # 获取特征重要性
# 通过二维表格形式显示
importances_df = pd.DataFrame()
importances_df['特征名称'] = features
importances_df['特征重要性'] = importances
importances_df.sort_values('特征重要性', ascending=False)

from sklearn.model_selection import cross_val_score
acc = cross_val_score(model, X, y, scoring='roc_auc', cv=5)
acc
acc.mean()

from sklearn.model_selection import GridSearchCV  # 网格搜索合适的超参数
# 指定参数k的范围
parameters = { 
'max_depth': [3, 5, 7, 9, 11]}
# 构建决策树分类器
model = DecisionTreeClassifier()  # 这里因为要进行参数调优，所以不需要传入固定的参数了
# 网格搜索
grid_search = GridSearchCV(model, parameters, scoring='roc_auc', cv=5)   # cv=5表示交叉验证5次，默认值为3；scoring='roc_auc'表示通过ROC曲线的AUC值来进行评分，默认通过准确度评分
grid_search.fit(X_train, y_train)
# 输出参数的最优值
grid_search.best_params_

from sklearn.model_selection import GridSearchCV
# 指定决策树分类器中各个参数的范围
parameters = { 
'max_depth': [5, 7, 9, 11, 13], 'criterion':['gini', 'entropy'], 'min_samples_split':[5, 7, 9, 11, 13, 15]}
# 构建决策树分类器
model = DecisionTreeClassifier()  # 这里因为要进行参数调优，所以不需要传入固定的参数了
# 网格搜索
grid_search = GridSearchCV(model, parameters, scoring='roc_auc', cv=5)
grid_search.fit(X_train, y_train)
# 获得参数的最优值
grid_search.best_params_
# 根据多参数调优的结果来重新搭建模型
model = DecisionTreeClassifier(criterion='entropy', max_depth=11, min_samples_split=13)
model.fit(X_train, y_train) 
# 查看整体预测准确度
y_pred = model.predict(X_test)
from sklearn.metrics import accuracy_score
score = accuracy_score(y_pred, y_test)
print(score)
# 查看新的AUC值
# 预测不违约&违约概率
y_pred_proba = model.predict_proba(X_test)
y_pred_proba[:,1]  # 如果想单纯的查看违约概率，即查看y_pred_proba的第二列
score = roc_auc_score(y_test, y_pred_proba[:,1])
print(score)