I'm doing uniqueness on factor loading matrix in a factor model.

$ y = \Lambda f + \epsilon$

where $ f \sim N(0,\Sigma)$ , $\epsilon \sim N(0,\Omega) $ and $\epsilon \perp f$.

It's well known that factor loading matrix $\Lambda$ is not identifiable.

My goal is to use a prior on $\Lambda$ such that it can be identified.

I know that there are many papers which have already work into this problem, but they are under the framework which assume $\Sigma$ and $\Omega$ are diagonal matrix.

However, I want to do that under the situation that $\Sigma$ and $\Omega$ are not diagonal matrix.

Has it already have paper on this part? Or is there any idea that I can achieve this?


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.