一、定义:
线性回归在假设特证满足线性关系,根据给定的训练数据训练一个模型,并用此模型进行预测。
二、代码:
import numpy as np
from matplotlib import pyplot as plt
X=np.array([2,3,4,5,6])
Y=2*X+np.random.normal(1,2,5)
plt.scatter(X,Y)
x_mean=np.mean(X)
y_mean=np.mean(Y)
n=0.0
d=0.0
for x,y in zip(X,Y):
n+=(x-x_mean)*(y-y_mean)
d+=(x-x_mean)**2
a=n/d
b=y_mean-a*x_mean
y_predict=[a*x+b for x in X]
plt.scatter(X,Y)
plt.plot(X,y_predict,color='r')
ss_residual=sum((y_predict-Y)**2)
ss_total=sum((Y-y_mean)**2)
score=1-ss_residual/ss_total
print(score)
n=5
betal_hat=a
se_model=np.sqrt(ss_residual/(n-2))
sss=np.sqrt(sum((X-x_mean)**2))
t_val=betal_hat/(se_model/sss)
from scipy.stats import t
p_val=2*(1-t.cdf(t_val,n-2))
print(p_val)
三、结果:
假设检验结果:
小于0.05,拒绝零假设,有线性关系。