1

I am studying econometrics, but due to my knowledge level I have to down to the basics. Can someone explain what $x^Tx$ does? I understand how calculations is done, but I don't see the intuitive $$\begin{pmatrix}1&1&1\\2&3&4\end{pmatrix}\begin{pmatrix}1&2\\1&3\\1&4\end{pmatrix} = \begin{pmatrix}3&9\\9&29 \end{pmatrix}$$

Could someone please take some time to explain what the result actually represent/mean?

Brad S.
  • 1,876
ALEXANDER
  • 2,099
  • try to look at this try to look at this

    http://math.stackexchange.com/questions/158219/is-a-matrix-multiplied-with-its-transpose-something-special

     – gbox Mar 25 '14 at 20:50
  • The example presented is incorrect. The product should be $\left(\begin{smallmatrix}5&7&9\7&10&13\9&13&17\end{smallmatrix}\right)$. – vadim123 Mar 25 '14 at 21:41
  • @vadim123 Ah, yes, you are correct but, I' m sure the OP meant it the other way around. I've submitted and edit. – Brad S. Mar 26 '14 at 16:57

1 Answers1

7

Based upon the form of your matrix $X$, and the fact that you called it $X$ and not $A$, I'm guessing your question comes from the matrix form of doing simple linear regression. So, you are trying to find the regression parameters, the slope and intercept of the fitted line. Recall that when you calculate those parameters, you need to know the sample size, $n$, and the sums, $\sum{x_i}$ and $\sum{x_i^2}$ among other things. Now, if you look at the entries in the result (referring to your example), you see that the result has the form...

$$\begin{pmatrix}n&\sum{x_i}\\\sum{x_i}&\sum{x_i^2}\end{pmatrix}$$

Hopefully, you can see how important these quantities are in the calculation of the linear regression parameters.

Brad S.
  • 1,876