Sign in
Microsoft.com
 
Microsoft
 
 
 

locked
Variational Mixture of Gaussians RRS feed

 > 

  • Question

  • Question
    Sign in to vote
    0
    Sign in to vote

    Hello,

    I was wondering if you can tell me how I need to put initial and consequent values for the approximate inference of Variational Mixture of Gaussians based on Bishop's Pattern Recognition and machine Learning (pages 478, 479) where he says: "Then in the subsequent variational equivalent of the M step, we keep these responsibilities fixed and use them to re-compute the variational distribution over the parameters using (10.57) and (10.59).". 

    I am not sure how exactly implement E and M steps regarding theses information.

    Thanks

    Wednesday, July 29, 2015 4:30 PM

All replies

  • Question
    Sign in to vote
    0
    Sign in to vote

    I don't have the book at hand, but here's how to write a Gaussian mixture in Infer.NET using VMP.

    -Y-

    Wednesday, July 29, 2015 5:52 PM
  • Question
    Sign in to vote
    0
    Sign in to vote

    Thank You ! The programming code doesn't matter to me. I don't know to reflect 10.64 and 10.65 into 10.57-10.59.

    Wednesday, July 29, 2015 6:10 PM
  • Question
    Sign in to vote
    0
    Sign in to vote
    This is not a forum for mathematical questions.  Try asking on Cross Validated.
    Wednesday, July 29, 2015 6:22 PM
    Owner