特征选择——变量重要性排序

1.利用决策树进行排序

步骤

• 导入数据；
• 建模
• 训练
• 获取变量重要性得分

from sklearn.tree import DecisionTreeClassifier
import numpy as np
from sklearn.datasets import load_wine
from model_selection import train_test_split
data = load_wine()  # 下载数据
'''data.data.shape	#数据变量
->	(178, 13)
	data.target # 类别
'''
X_train, X_test, Y_train, Y_test = train_test_split(data.data,data.target,random_state=33)
model = DecisionTreeClassifier(random_state=2)
model.fit(X_train,Y_train)
features_importance = model.feature_importances_
# 重要性排序降序
indices = np.argsort(features_importance)[::-1])
# 返回排序后的下标,并取前几个重要的特征
importance = indices[:3]
print(X_train[:,importance])

2. 利用随机森林进行变量选择

from sklearn.ensemble import RandomForestClassifier as RF
import numpy as np
from sklearn.datasets import load_wine
from model_selection import train_test_split
data = load_wine()  # 下载数据
'''data.data.shape	#数据变量
->	(178, 13)
	data.target # 类别
'''
X_train, X_test, Y_train, Y_test = train_test_split