Take the 2-minute tour ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

I have a deterministic circulant matrix $R$ and a random diagonal matrix $X$ where all elements are IID and positive. I need to determine the expected inverse of $R+X$, that is: Evaluate, in closed, form $E[(R+X)^{-1}]$.

I am mostly interested in an inverse chi-square distribution for each element of $X$ with 2 degrees of freedom (so $E[X]$ does not exist). However, I would appreciate any help you can provide for any distribution.

By simple playing around in matlab, the solution appears to be of the form $(R+\lambda I)^{-1}$, where $\lambda$ depends on the distribution of $X$. This holds for all distributions I have tried (Gauss, chi-2, inverse-chi-2, a couple of discrete ones...)

One simplification that I am particularly interested is: Let the dimensions of $R$ and $X$ tend to infinity, and let $R$ have an eigenvalue spectrum $R(\omega)=1, |\omega|<\delta 2\pi$ and $R(\omega)=0, |\omega|\geq\delta 2\pi$ for some $0<\delta <1.$

share|improve this question
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.