Suppose X is distributed as multivariate Gaussian centered at 0 with dxd covariance matrix sigma. Let v be some d x 1 vector. What is P(X=x|v'x=a)? Specifically, I want to know the technique for computing the normalization factor.

Update 16/09 well-informed fellows on math.stackoverflow show how to do it using Dirac measure trick

asked Sep 07 '10 at 14:35

Yaroslav%20Bulatov's gravatar image

Yaroslav Bulatov

edited Sep 16 '10 at 20:31

Be the first one to answer this question!
toggle preview

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.