——王衡军编《机器学习》P78-79
1、用sklearn.limear_model包中的LinearRegression对表3-1所示的示例进行线性回归实验,比较结果。
实验代码:
from sklearn import linear_model # 导入线性模型
import matplotlib.pyplot as plt#用于作图
import numpy as np#用于创建向量
reg=linear_model.LinearRegression(fit_intercept=True,normalize=False)
x = [[15], [20], [25], [30], [35],[40]]
y = [136, 140, 155, 160, 157, 175]
reg.fit(x,y)
k=reg.coef_#获取斜率w1,w2,w3,...,wn
b=reg.intercept_#获取截距w0
x0=np.arange(15,40)
y0=k*x0+b
plt.scatter(x,y)
plt.plot(x0,y0)
print("k=",k,"b",b ,"\n")
new_tempera = [18]
new_tempera = (np.mat(new_tempera)).T
pro_num = k*new_tempera+b
print(pro_num)
new_tempera = [22]
new_tempera = (np.mat(new_tempera)).T
pro_num = k*new_tempera+b
print(pro_num)
new_tempera = [33]
new_tempera = (np.mat(new_tempera)).T
pro_num = k*new_tempera+b
print(pro_num)
5、实现岭回归的最小二乘法求解算法,并进行实