先来个前置知识：
E
[
X
Y
∣
X
]
=
X
⋅
E
[
Y
∣
X
]
(1)
E[XY|X]=X\cdot E[Y|X]\tag{1}
E[XY∣X]=X⋅E[Y∣X](1)
E
[
E
(
X
∣
Y
)
]
=
E
[
X
]
(2)
E[E(X|Y)]=E[X]\tag{2}
E[E(X∣Y)]=E[X](2)
Cov ( X , Y ) = E [ X Y ] − E [ X ] E [ Y ] (3) \text{Cov}(X,Y)=E[XY]-E[X]E[Y]\tag{3} Cov(X,Y)=E[XY]−E[X]E[Y](3)
Cov ( X , E [ Y ∣ X ] ) = E [ X ⋅ E ( Y ∣ X ) ] − E [ X ] ⋅ E [ E ( Y ∣ X ) ] = E [ E ( X Y ∣ X ) ] − E [ X ] ⋅ E [ E ( Y ∣ X ) ] = E [ E ( X Y ∣ X ) ] − E [ X ] E [ Y ] = E [ X Y ] − E [ X ] E [ Y ] = Cov ( X , Y ) (4) \begin{align*}\text{Cov}(X, E[Y|X])&=E[X\cdot E(Y|X)]-E[X]\cdot E[E(Y|X)]\\&=E[E(XY|X)]-E[X]\cdot E[E(Y|X)]\\&=E[E(XY|X)] - E[X]E[Y]\\&=E[XY]-E[X]E[Y]=\text{Cov}(X,Y)\end{align*}\tag{4} Cov(X,E[Y∣X])=E[X⋅E(Y∣X)]−E[X]⋅E[E(Y∣X)]=E[E(XY∣X)]−E[X]⋅E[E(Y∣X)]=E[E(XY∣X)]−E[X]E[Y]=E[XY]−E[X]E[Y]=Cov(X,Y)(4)
证明来源：Show that Cov(X,Y)=Cov(X,E(Y|X)).
设线性回归模型：
y
=
α
+
β
x
+
μ
(1*)
y=\alpha + \beta x + \mu \tag{1*}
y=α+βx+μ(1*)
如果零均值假设
E
(
u
∣
x
)
=
0
E(u|x)=0
E(u∣x)=0成立，则
E ( y ∣ x ) = α + β x (2*) E(y|x)=\alpha + \beta x \tag{2*} E(y∣x)=α+βx(2*)
对 x x x求协方差：
Cov x [ x , E ( y ∣ x ) ] = Cov x ( x , α ) + β Cov x ( x , x ) (3*) \text{Cov}_x[x,E(y|x)]=\text{Cov}_x(x,\alpha)+\beta \text{Cov}_x(x,x)\tag{3*} Covx[x,E(y∣x)]=Covx(x,α)+βCovx(x,x)(3*)
变量和常量的协方差为0，变量与自身的协方差即变量的方差：
Cov
(
x
,
y
)
=
β
Var
(
x
)
(4*)
\text{Cov}(x,y)=\beta \text{Var}(x)\tag{4*}
Cov(x,y)=βVar(x)(4*)
通过简单的变换可以获得：
β = Cov ( x , y ) Var ( x ) (5*) \beta = \frac{\text{Cov}(x,y)}{\text{Var}(x)}\tag{5*} β=Var(x)Cov(x,y)(5*)
使用样本矩来近似估计总体矩：
β ^ = ∑ i = 1 n ( x − x ˉ ) ( y − y ˉ ) ∑ i = 1 n ( x − x ˉ ) 2 (6*) \hat \beta = \frac{\sum\limits_{i=1}^n(x-\bar{x})(y-\bar{y})}{\sum\limits_{i=1}^n(x-\bar{x})^2}\tag{6*} β^=i=1∑n(x−xˉ)2i=1∑n(x−xˉ)(y−yˉ)(6*)
9151
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
