[回归算法]常用功能实现

前言：回归任务是监督式机器学习中的最主要的一种类别，与分类不同的是，其目标一般为连续型数值。在sklearn中提供了大量的用于回归的算法。

本文汇总了常见的9种回归算法，方便快速查询使用。（本文使用糖尿病检测数据集，使用MSE作为回归评价指标，以下为MSE得分）

1.线性回归：3424

2.岭回归：3379

3.套索回归：3787

4.弹性网络：4666

5.支持向量机(多项式核)：4267

6.K近邻：4243

7.决策树：7762

8.随机森林：3925

9.梯度提升树：4112

import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt
import sklearn
from sklearn import datasets

from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score
from sklearn.metrics import explained_variance_score
from sklearn.metrics import mean_squared_error

from sklearn.linear_model import LinearRegression
from sklearn.linear_model import Ridge
from sklearn.linear_model import Lasso
from sklearn.linear_model import ElasticNet

from sklearn.svm import SVR
from sklearn.neighbors import KNeighborsRegressor
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.ensemble import GradientBoostingRegressor


diabetes = sklearn.datasets.load_diabetes()
x,y = diabetes.data,diabetes.target
x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.2,random_state=0)
res = []

#1. 线性回归
regr = LinearRegression()
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#2. 岭回归
regr = Ridge()
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#3. 套索回归
regr = Lasso()
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#4. 弹性网络
regr = ElasticNet(alpha=0.1,l1_ratio=0.5)
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#5. SVR
regr = SVR(gamma='scale',kernel="poly")#linear rbf
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#6. K近邻回归
regr = KNeighborsRegressor(weights="uniform")
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#7. 决策树回归
regr =DecisionTreeRegressor()
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#8. 随机森林回归
regr = RandomForestRegressor(n_estimators=100)
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#9. 梯度提升树回归
regr = GradientBoostingRegressor()
regr.fit(x_train, y_train)
cross_score = cross_val_score(regr, x_train, y_train, cv=3, scoring="neg_mean_squared_error")
print(-cross_score)
y_predict = regr.predict(x_test)
score = mean_squared_error(y_test,y_predict)
res.append(score)

#10. 得分比较
print(res)