1

I have the following likelihood function:

enter image description here

I'm given this information about the $\Omega$ matrix ($\boldsymbol{1}$ is a $T \times 1$ vector of ones): enter image description here

I would like to be able to show that the likelihood function can be rewritten to this: enter image description here

I think there is a misprint in the article, hence the screen images from the article I attached are wrong, meaning that $\Omega^{-1}$ is supposed to be $$\Omega^{-1}=\frac{1}{\sigma^2_{\varepsilon}+T\sigma^2_{c}}\boldsymbol{1}(\boldsymbol{1}'\boldsymbol{1})^{-1}\boldsymbol{1}'+\frac{1}{\sigma^2_{\varepsilon}}(I_t-\boldsymbol{1}(\boldsymbol{1}'\boldsymbol{1})^{-1}\boldsymbol{1}')$$

and not as stated above.

So far I've inserted the expression for $\Omega^{-1}$ in the top log-likelihood function. However, especially the last expression of the rewritten log-likelihood function bothers me, i.e. this part:

\begin{equation*} \begin{split} ...& &+\frac{1}{2}\frac{1}{\sigma^2_{\varepsilon}}\frac{1}{N}\sum_{i=1}^{N} \left(Y_i-\boldsymbol{1}\bar{Y}_i-\left(X_i-\boldsymbol{1}\bar{X}_i\right)\beta\right)'\left(Y_i-\boldsymbol{1}\bar{Y}_i-\left(X_i-\boldsymbol{1}\bar{X}_i\right)\beta\right). \end{split} \end{equation*}

Does anyone see the solution?

I also have this lemma, but not sure if it is to be used in the rewritting of the likelihood function: enter image description here

Sunv
  • 321
  • 3
  • 12
  • @pbr142 is this something you could help me explain? – Sunv Apr 08 '14 at 14:02
  • 1
    This is basic linear algebra, notably block matrix inversion, and more specifically rank one update. – nicolas Apr 09 '14 at 08:26
  • @nicolas Thanks for your answer. Could you be more specific as to how I should obtain the rewritten likelihood function? – Sunv Apr 09 '14 at 08:31
  • I have updated the question for more clarity – Sunv Apr 09 '14 at 08:35
  • 1
    read about block inversion/rank one update, and apply... :) Look at the Horn Johnson or Golub – nicolas Apr 09 '14 at 10:19
  • @nicolas hmm, doesn't necessarily need the involvement of block inversion/rank one update (I'm out of my comfort zone so bear with me if I'm not correct here). I have added a lemma that is used to show how to write $\Omega$. I hope this helps. – Sunv Apr 09 '14 at 10:52
  • @pbr142 Do you have any inputs? – Sunv Apr 09 '14 at 10:53
  • 1
    what you added is a special case with block diagonal matrix. I know it's painful, but you should take a few days to get into the comfort zone, this basic linear algebra will help you long time – nicolas Apr 09 '14 at 11:22
  • @nicolas I'm starting to realize you might be right! I have worked with simple linear algebra, so the subject is not completely new to me. Do you have any ideas as to what I should read/understand in order to do the rewritting myself? – Sunv Apr 09 '14 at 11:32
  • I was in the same situation ;) As a cure, I recently swallowed Horn, Matrix Analysis up to the middle of the book. The other reference is Golub Matrix computation, did not try it. I think Linear Algebra is never a let down. it's a friend for life, never a waste to master. You can try to jump to the relevant section of block manipulation (in chapter of Horn) and work any concept backward, I think it is very self contained. – nicolas Apr 09 '14 at 12:31

0 Answers0