MMSE

Dear all,

Using Matlab I was trying to find unknown vector 'omega' using Minimum
Mean Square Error method given by

omega_cap=C_omegaY*(C_YY)^(-1)*Y

in the above equation the only known vector is Y.

where C_omegaY = cross correlation between omega and Y
 and  C_YY     = auto correlation of Y

when my task is to estimate the vector omega, how could I find the cross
correlation between Y and unknown vecotor omega.

Do I have to assume some values for vector omega in the begining???? If
so, what values would be appropriate????

Thanks in advance

On Mar 11, 5:55 am, "aamer" <raqeeb...@yahoo.com> wrote:
> Dear all,
>
> Using Matlab I was trying to find unknown vector 'omega' using Minimum
> Mean Square Error method given by
>
> omega_cap=C_omegaY*(C_YY)^(-1)*Y
>
> in the above equation the only known vector is Y.
>
> where C_omegaY = cross correlation between omega and Y
>  and  C_YY     = auto correlation of Y
>
> when my task is to estimate the vector omega, how could I find the cross
> correlation between Y and unknown vecotor omega.
>
> Do I have to assume some values for vector omega in the begining???? If
> so, what values would be appropriate????
>
> Thanks in advance

Are you saying that you observe a vector y, and you want to
know about another vector omega.  But you don't know how
they are related.  Can you tell me how this makes sense?

Why don't you explain what you have, what you want, and
why you want to have what you want? :-)

In the above setting usually you are given some statistical
idea of how omega and y are related.  If you don't have any
such knowledge, why are you trying to do what you are
trying to do?

Did you pick the above our of some textbook or paper?

Julius