0
$\begingroup$

As part of my research, I would like to apply the Metropolis-Hastings in order to sample from some posterior distribution. More precisely, the data comes from a multivariate normal distribution in the presence of censorship and the prior has normal-inverse-wishart distribution (the covariance matrix has inverse-wishart distribution and the mean conditioned on the covariance matrix has normal distribution). One possible proposal distribution for the Metropolis-Hastings algorithm consists in using the prior distributions itself since it is easy to sample from it. However the acceptance ratio is generally very low and the algorithm evolves quite slowly. Having said that, I would like to know if anyone could provide me some guidance on the subject, that is to say, I would be interested in knowing if the choice for the prior distribution is acceptable as well as whether there exists another proposal distribution which is more efficient.

Any help is appreciated.

The same question has been asked here.

Improve this question
$\endgroup$
3
  • $\begingroup$ I've normally seen it called "censoring" rather than "censorship." $\endgroup$
    – Michael Hardy
    Commented Aug 28 at 16:16
  • $\begingroup$ Since you've gotten tons of relevant info there and none here, does it make sense to delete this question? $\endgroup$
    – John Madden
    Commented Aug 31 at 16:58
  • $\begingroup$ Although they have provided many contributions, none of them fully answer my question. $\endgroup$
    – learner123
    Commented Sep 4 at 0:20

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .