MAP Estimation of Covariance Matrix of a Multivariate Normal Distribution

343 Views Asked by At

I have a general prior multivariate normal distribution and I want to update it with new samples which are more local for my case. I want to do it with MAP estimation. With MAP estimation it is possible to find a relation for new mean vector, however I could not find a general expression for the covariance estimation. Is there a general way of estimating covariance matrix in the framework of MAP estimation, or I need to have some assumptions and specific cases for estimating it?