实现gbdt算法

integr/gbdt_train.py

@@ -3,12 +3,12 @@
 #-*- coding:utf-8 -*-
  
 import numpy as np
-from sklearn.ensemble import GradientBoostingClassifier
+from sklearn.ensemble import GradientBoostingClassifier,GradientBoostingRegressor
 from sklearn.model_selection import train_test_split
-from sklearn.datasets import load_wine
+from sklearn.datasets import load_wine,load_boston
 wine = load_wine()
 Xtrain, Xtest, Ytrain, Ytest = train_test_split(wine.data,wine.target,test_size=0.3)
-#print (X,y)
+print(Ytrain)
  
 #默认参数
 #Accuracy : 0.9856
@@ -16,5 +16,22 @@ Xtrain, Xtest, Ytrain, Ytest = train_test_split(wine.data,wine.target,test_size=
 gbm1 = GradientBoostingClassifier( n_estimators=500,max_depth=10,max_features='sqrt', random_state=10)
 gbm1.fit(Xtrain,Ytrain)
 print("gbdt1",gbm1.score(Xtest,Ytest))
- 
- 
+
+boston = load_boston()
+Xtrain, Xtest, Ytrain, Ytest = train_test_split(boston.data, boston.target, test_size=0.3)
+print(Ytrain)
+
+# 默认参数
+# Accuracy : 0.9856
+# AUC Score (Train): 0.862264
+gbm2 = GradientBoostingRegressor(n_estimators=5, max_depth=5, max_features='sqrt', random_state=10)
+gbm2.fit(Xtrain, Ytrain) # 分数越高越好
+print("gbdt1", gbm2.score(Xtest, Ytest))
+
+gbm2 = GradientBoostingRegressor(n_estimators=50, max_depth=5, max_features='sqrt', random_state=10)
+gbm2.fit(Xtrain, Ytrain)
+print("gbdt2", gbm2.score(Xtest, Ytest))
+
+gbdt3 = GradientBoostingRegressor(n_estimators=150, max_depth=5, max_features='sqrt', random_state=10)
+gbdt3.fit(Xtrain, Ytrain)
+print("gbdt3", gbdt3.score(Xtest, Ytest))