Gaussian Process Regression Networks

All replies

• 

  

  0

  Sign in to vote

  Hi Eric

  The Infer.NET team does not have that code. It looks like the authors built a custom factor and operator to do that. You should contact the authors directly, and perhaps encourage them to post the answer here.

  John

  Friday, July 26, 2013 12:36 PM

  Owner
• 

  

  0

  Sign in to vote

  Hi Eric

  The Infer.NET team does not have that code. It looks like the authors built a custom factor and operator to do that. You should contact the authors directly, and perhaps encourage them to post the answer here.

  John

  Are any samples available which Gaussian Process Regression rather than classification.  Perhaps that go through the process of optimizing hyper parameters and such as well.

  Friday, July 26, 2013 6:54 PM
• 

  

  0

  Sign in to vote

   I don't think we have a sample, but you can remove the ' > 0' at the end of the definition of y[j] in the example, and the classifier becomes a regressor. 

  Sunday, July 28, 2013 5:48 PM
• 

  

  0

  Sign in to vote

  Also see forum thread http://social.microsoft.com/Forums/en-US/1be040b0-5466-487b-93e8-d2058241a33b/matrix-is-not-positive-definite-migrated-from-communityresearchmicrosoftcom

  • Proposed as answer by John GuiverMicrosoft employee, Owner Wednesday, August 14, 2013 7:40 AM

  Tuesday, July 30, 2013 3:07 PM

  Owner
• 

  

  0

  Sign in to vote

  I sent this to Eric personally a couple of months ago but in case anyone else is interested, the code is up at

  https://github.com/davidaknowles/gprn

  Best,

  David

  • Marked as answer by John GuiverMicrosoft employee, Owner Friday, December 20, 2013 11:05 AM

  Wednesday, October 30, 2013 11:31 PM