Multivariate normal distribution

[SYP]

Hi All,

I am trying to solve this brainier, but with little success. First of all apologies that it's coming as a picture, I couldn't make Latex work.

Each vector has the multivariate normal distribution, X bar is the mean vector of all vectors. The solution in my opinion involves eigenvalues and eigenvectors, but how ?

Thank you in advance !

 

[fed2]

Im getting the vibe that you can solve it by setting v to be the 'stack' of the X_i, then setting A to give the arithmetic mean of appropriate elements of v? Why do you say eigen-vectors? Maybe, can't say i do much eigenvalues these days.