Sign in
Microsoft.com
 
Microsoft
 
 
 

locked
MIxture Model using constraints RRS feed

 > 

  • Question

  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi everyone,

    I am wondering if it is possible to infer a mixture model using constraints?

    To get more specific, I am interessted in using an Exponential-Uniform or Gamma-Uniform model and constrain the Exponential or Gamma distribution between 0 and 1.

    Is there a way to implement models like these?

    Thanks!

    Wednesday, August 12, 2015 10:28 AM

All replies

  • Question
    Sign in to vote
    0
    Sign in to vote
    If you are trying to learn the parameters of a truncated distribution then you would need to write a custom factor to take account of the modified normalizing constant.  These are not currently provided by Infer.NET.
    Wednesday, August 12, 2015 12:42 PM
    Owner