2
$\begingroup$

All is in matrix notation!
Define the error sum squares as:
$SS_e(\beta)=(y-X\beta)'(y-X\beta)=y'y-2\beta'X'y+\beta'X'X\beta$

For the OLS estimator $\hat{\beta}$, show that:
$SS_e(\beta)=SS_{Res}+(\beta-\hat{\beta})'X'X(\beta-\hat{\beta})$

So we have: $SS_{Res}=(y-X\hat{\beta})'(y-X\hat{\beta})=y'y-y'X\hat{\beta}-\hat{\beta}'X'y+\hat{\beta}'X'X\hat{\beta}=y'y-2\hat{\beta}'X'y+\hat{\beta}'X'X\hat{\beta}$

and $(\beta-\hat{\beta})'X'X(\beta-\hat{\beta})=\beta'X'X\beta-\beta'X'X\hat{\beta}-\hat{\beta}'X'X\beta+\hat{\beta}'X'X\hat{\beta}=\beta'X'X\beta-2\hat{\beta}'X'X\beta+\hat{\beta}'X'X\hat{\beta}$

Together we get:
$y'y-2(\hat{\beta}'X'y+\hat{\beta}'X'X\beta-\hat{\beta}'X'X\hat{\beta})+\beta'X'X\beta$

Here is where I run into trouble, I can't manage to show that $\hat{\beta}'X'y+\hat{\beta}'X'X\beta-\hat{\beta}'X'X\hat{\beta}=\beta'X'y$

Could really need some help, thx.

$\endgroup$
  • $\begingroup$ Ok, think I solved it. $\endgroup$ – Something Mar 2 '18 at 10:27
1
$\begingroup$

$\hat\beta'X'y=y'X(X'X)^{-1}X'y=y'Hy=y'\hat y=y'(y+e)=y'y+y'e=y'y$
$\hat\beta'X'X\beta=y'X(X'X)^{-1}X'X\beta=y'X\beta=\beta'X'y$
$\hat\beta'X'X\hat\beta=y'X(X'X)^{-1}X'X(X'X)^{-1}X'y=y'Hy=y'y$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.