I have a set of random vectors $Y_i$ and their correlation matrix $C_{i,j}$. Each vector can be thought of as a sum of two uncorrelated vectors $Y_i=A_iX+B_iY$, where $X,Y$ are the same vectors for each $Y_i$ and $A_i,B_i$ are some scalar coefficients. One can assume that $X$ and $Y$ are uncorrelated. The goal is to find all $A_i$, $B_i$ (for every $i$).

So far I was thinking to use the cholesky decomposition $D$ of the correlation matrix $C$ so that $(A_iX+B_iY)D=C$, and then solve the system for A and B. The problem is, however, that I do have two more variables than equations. So, the question is: how to use the fact that X and Y are uncorrelated to exclude them from the picture and just find A and B? Or is there other method to solve the problem? Thanks to everyone!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.