The Road to Machine Learning: Python Ensemble Regression Model Random Forest Regression RandomForestRegressor Extreme Random Forest Regression ExtraTreesRegressor GradientBoostingRegressor Regression Predicting Boston Housing Prices

python3 learn machine learning api

Three ensemble regression models were used

git: https://github.com/linyi0604/MachineLearning

Code:

  1 from sklearn.datasets import load_boston
  2 from sklearn.cross_validation import train_test_split
  3 from sklearn.preprocessing import StandardScaler
  4 from sklearn.ensemble import RandomForestRegressor, ExtraTreesRegressor, GradientBoostingRegressor
  5 from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error
  6 import numpy as np
  7 
  8 '''
  9 随机森林回归
 10 极端随机森林回归
 11 Gradient boosting regression
 12  
13  Usually the ensemble model can achieve very good performance
 14  ''' 
15  
16  # 1 Prepare data 
17 #Read  the housing price information in the Boston area 18 boston = load_boston()
 19 #View data description 20 # print ( boston.DESCR ) # A total of 506 pieces of housing price information in the Boston area, each with 13 numerical feature descriptions and the target housing price 21 # Check the difference between the data 22 # print("Maximum housing price:", np.max(boston.target)) # 50 23 # print("Minimum house price:",np.min(boston.target)) # 5 24 # print("Average house price:", np.mean(boston.target)) # 22.532806324110677 25 26 x = boston.data

 
 
 
 
 
 
 
27 y = boston.target
 28  
29  # 2 split training data and test data 
30  # randomly sample 25% as test 75% as training 
31 x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25, random_state=33 )
 32  
33  # 3 Normalize training data and test data 
34 ss_x = StandardScaler()
 35 x_train = ss_x.fit_transform(x_train)
 36 x_test = ss_x.transform(x_test)
 37  
38 ss_y = StandardScaler()
 39 y_train = ss_y.fit_transform (y_train.reshape(-1, 1 ))
 40y_test = ss_y.transform(y_test.reshape(-1, 1 ))
 41  
42  # 4 Three integrated regression models for training and prediction 
43  #Random forest regression 
44 rfr = RandomForestRegressor()
 45  #training 46 rfr.fit 
(x_train, y_train)
 47 #Predict save the prediction result 48 rfr_y_predict = rfr.predict(x_test)
 49 50 #Extreme random forest regression 51 etr = ExtraTreesRegressor()
 52 #train 53 etr.fit (x_train, y_train)
 54 #Predict save the prediction result 55 etr_y_predict =  
 
 
 
  
rfr.predict(x_test)
 56  
57  #gradient boosting regression 
58 gbr = GradientBoostingRegressor()
 59  #training 60 gbr.fit 
(x_train, y_train)
 61 #prediction save prediction result 62 gbr_y_predict = rfr.predict(x_test)
 63 64 # 5 model Evaluation 65 #Random forest regression model evaluation 66 print ( " The default evaluation value of random forest regression is: " , rfr.score(x_test, y_test))
 67 print ( " The R_squared value of random forest regression is: " , r2_score(y_test, rfr_y_predict))
 68  
 
 
 
   print ( " The mean squared error of random forest regression is: " , mean_squared_error(ss_y.inverse_transform(y_test),
 69                                            ss_y.inverse_transform(rfr_y_predict)))
 70  print ( " The mean absolute error of random forest regression is: " , mean_absolute_error(ss_y .inverse_transform(y_test),
 71                                               ss_y.inverse_transform(rfr_y_predict)))
 72  
73  #Extreme random forest regression model evaluation 
74  print ( " The default evaluation value of extreme random forest regression is: " , etr.score(x_test, y_test))
 75  print ( "The R_squared value of extreme random forest regression is: " , r2_score(y_test, gbr_y_predict))
 76  print ( " The mean squared error of extreme random forest regression is: " , mean_squared_error(ss_y.inverse_transform(y_test),
 77                                              ss_y.inverse_transform(gbr_y_predict) ))
 78  print ( " The mean absolute error of extreme random forest regression is: " , mean_absolute_error(ss_y.inverse_transform(y_test),
 79                                                 ss_y.inverse_transform(gbr_y_predict)))
 80  
81  #gradient boosting regression model evaluation 
82  print ( " gradient boosting The default evaluation value for regression regression is: ", gbr.score(x_test, y_test))
 83  print ( " The R_squared value of gradient boosting regression is: " , r2_score(y_test, etr_y_predict))
 84  print ( " The mean squared error of gradient boosting regression is: " , mean_squared_error( ss_y.inverse_transform(y_test),
 85                                              ss_y.inverse_transform(etr_y_predict)))
 86  print ( " The mean absolute error of gradient boosted regression is: " , mean_absolute_error(ss_y.inverse_transform(y_test),
 87                                                 ss_y.inverse_transform(etr_y_predict)))
 88  
89  ''' 
90 The default evaluation value for random forest regression is: 0.8391590262557747
 91  The R_squared value for random forest regression is: 0.8391590262557747
 92  The mean squared error for random forest regression is: 12.471817322834646
 93  The mean absolute error for random forest regression is
 : 2.4255118110236227 94  
The evaluation value is: 0.783339502805047
 96 The R_squared value of the extreme random forest regression is: 0.8391590262557747
 97 The mean squared error of the extreme random forest regression is: 12.471817322834646
 98 The mean absolute error of the extreme random forest regression is: 2.4255118110236227
 The default value of the Boostinggress regression is 100 0.8431187344932869
 101 The R_squared value of the GradientBoostingRegressor regression is: 0.8391590262557747
 102 The mean squared error of the GradientBoostingRegressor regression is: 12.471817322834646
 103     
    The mean absolute error of GradientBoostingRegressor regression is: 2.4255118110236227
 104  '''