Python3 Learning using the API
Linear regression, and stochastic parametric regression
Git:https://github.com/linyi0604/machinelearning
1 fromSklearn.datasetsImportLoad_boston2 fromSklearn.cross_validationImportTrain_test_split3 fromSklearn.preprocessingImportStandardscaler4 fromSklearn.linear_modelImportlinearregression, Sgdregressor5 fromSklearn.metricsImportR2_score, Mean_squared_error, Mean_absolute_error6 ImportNumPy as NP7 8 #1 Preparing Data9 #Read the Boston area rate informationTenBoston =Load_boston () One #View Data Description A #Print (Boston. DESCR) # A total of 506 Boston area rate information, each 13 numerical description and target rate - #view differences in Data - #print ("max rate:", Np.max (Boston.target)) # the #print ("Minimum rate:", Np.min (Boston.target)) # 5 - #print ("Average price:", Np.mean (Boston.target)) # 22.532806324110677 - -x =Boston.data +y =Boston.target - + #2 split training data and test data A #random sample 25% as test 75% as training atX_train, X_test, y_train, y_test = Train_test_split (x, Y, test_size=0.25, random_state=33) - - - #3 Standardized processing of training data and test data -Ss_x =Standardscaler () -X_train =ss_x.fit_transform (X_train) inX_test =ss_x.transform (x_test) - toSs_y =Standardscaler () +Y_train = Ss_y.fit_transform (Y_train.reshape (-1, 1)) -Y_test = Ss_y.transform (Y_test.reshape (-1, 1)) the * #4 Training and forecasting using two linear regression models $ #Initialize linearregression linear regression modelPanax NotoginsengLR =linearregression () - #Training the Lr.fit (X_train, Y_train) + #forecast Save Forecast results ALr_y_predict =lr.predict (x_test) the + #initializing Sgdrregressor stochastic gradient regression model -SGDR =sgdregressor () $ #Training $ Sgdr.fit (X_train, Y_train) - #forecast Save Forecast results -Sgdr_y_predict =sgdr.predict (x_test) the - #5 Model EvaluationWuyi #evaluation of the linear model theLr_score =Lr.score (x_test, y_test) - Print("the default evaluation value for linear is:", Lr_score) Wulr_r_squared =R2_score (y_test, lr_y_predict) - Print("the r_squared value of the linear is:", lr_r_squared) AboutLr_mse =Mean_squared_error (Ss_y.inverse_transform (y_test), Ss_y.inverse_transform (lr_y_predict)) $ Print("the mean square error of the linear is:", Lr_mse) -Lr_mae =Mean_absolute_error (Ss_y.inverse_transform (y_test), Ss_y.inverse_transform (lr_y_predict)) - Print("the average absolute error of the linear is:", Lr_mae) - A #evaluation of the SGD model +Sgdr_score =Sgdr.score (x_test, y_test) the Print("the default evaluation value for SGD is:", Sgdr_score) -sgdr_r_squared =R2_score (y_test, sgdr_y_predict) $ Print("the r_squared value for SGD is:", sgdr_r_squared) theSgdr_mse =Mean_squared_error (Ss_y.inverse_transform (y_test), Ss_y.inverse_transform (sgdr_y_predict)) the Print("the mean square error of SGD is:", Sgdr_mse) theSgdr_mae =Mean_absolute_error (Ss_y.inverse_transform (y_test), Ss_y.inverse_transform (sgdr_y_predict)) the Print("the average absolute error of SGD is:", Sgdr_mae) - in " " the the default evaluation value for linear is: 0.6763403830998702 the the r_squared value for linear is: 0.6763403830998701 About the mean square error of linear is: 25.09698569206773 the the average absolute error of the linear is: 3.5261239963985433 the the the default evaluation value for SGD is: 0.659795654161198 + the r_squared value for SGD is: 0.659795654161198 - the mean square error of SGD is: 26.379885392159494 the the average absolute error of SGD is: 3.5094445431026413Bayi " "
Machine learning path: Python linear regression linearregression, stochastic parametric regression sgdregressor forecast Boston rates