我在测试下面的样本代码。所有分类结果都相当不错且合理（80%或更高）。所有回归结果却非常糟糕且异常（约20%）。为什么会这样？我一定是做错了什么，但我不清楚问题出在哪里。

import pandas as pdimport numpy as np#reading the datasetdf=pd.read_csv("C:\\my_path\\train.csv")#filling missing valuesdf['Gender'].fillna('Male', inplace=True)df.fillna(0)df.Loan_Status.replace(('Y', 'N'), (1, 0), inplace=True)#split dataset into train and testfrom sklearn.model_selection import train_test_splittrain, test = train_test_split(df, test_size=0.3, random_state=0)x_train=train.drop(['Loan_Status','Loan_ID'],axis=1)y_train=train['Loan_Status']x_test=test.drop(['Loan_Status','Loan_ID'],axis=1)y_test=test['Loan_Status']#create dummiesx_train=pd.get_dummies(x_train)x_test=pd.get_dummies(x_test)# Baggin Classifierfrom sklearn.ensemble import BaggingClassifierfrom sklearn import treemodel = BaggingClassifier(tree.DecisionTreeClassifier(random_state=1))model.fit(x_train, y_train)model.score(x_test,y_test)# Bagging Regressorfrom sklearn.ensemble import BaggingRegressormodel = BaggingRegressor(tree.DecisionTreeRegressor(random_state=1))model.fit(x_train, y_train)model.score(x_test,y_test)# AdaBoostClassifierfrom sklearn.ensemble import AdaBoostClassifiermodel = AdaBoostClassifier(random_state=1)model.fit(x_train, y_train)model.score(x_test,y_test)# AdaBoostRegressorfrom sklearn.ensemble import AdaBoostRegressormodel = AdaBoostRegressor()model.fit(x_train, y_train)model.score(x_test,y_test)# GradientBoostingClassifierfrom sklearn.ensemble import GradientBoostingClassifiermodel= GradientBoostingClassifier(learning_rate=0.01,random_state=1)model.fit(x_train, y_train)model.score(x_test,y_test)# GradientBoostingRegressorfrom sklearn.ensemble import GradientBoostingRegressormodel= GradientBoostingRegressor()model.fit(x_train, y_train)model.score(x_test,y_test)# XGBClassifierimport xgboost as xgbmodel=xgb.XGBClassifier(random_state=1,learning_rate=0.01)model.fit(x_train, y_train)model.score(x_test,y_test)# XGBRegressorimport xgboost as xgbmodel=xgb.XGBRegressor()model.fit(x_train, y_train)model.score(x_test,y_test)

样本数据来自下面的链接。

https://www.kaggle.com/wendykan/lending-club-loan-data

最后，这里是我看到的一些小样本。

# Bagging Regressorfrom sklearn.ensemble import BaggingRegressorregressor = BaggingRegressor()regressor.fit(x_train,y_train)accuracy = regressor.score(x_test,y_test)print(accuracy*100,'%')# result:13.022388059701505 %from sklearn.linear_model import LinearRegressionregressor = LinearRegression()regressor.fit(x_train,y_train)accuracy = regressor.score(x_test,y_test)print(accuracy*100,'%')# result:29.836209522493196 %

回答：

回归和分类是两种不同的任务。从你的代码来看，你似乎在用与分类器相同的数据来拟合回归器。基本上，回归器试图找到一个函数，以最佳方式猜测基于输入的输出数字。因此，目标值应该是连续空间的数字，而非类别。例如，你可能想根据借款人借出的金额来预测其收入。

查看这篇Medium文章以了解更多关于回归和分类之间的区别。