Kernel BP

[Scott Richardson]

I was reading Song et al.'s Kernel Belief Propagation paper [SonGreBicLowGue11] and noticed on the webpage of one of the authors that a "fast implementation for use in applications and benchmarks is provided within Graphlab". I've seen that GraphLab 1.0 contains a Non-parametric belief propagation algorithm [SudIhlFreWil03], however that seems to actually be an older algorithm referred to by the authors as Gaussian Mixture Belief Propagation.

Does anyone know if Kernel BP is part of GraphLab? and if so, is it part of GraphLab 1.0 or 2.1? 

thanks
Scott

[Danny Bickson]

Hi Scott!
I did some digging - previously we had a KBP graphlab code for
version 1 - which was later removed from the repository. The code was
implemented by Le Song (now in Gatech). Le is cc-ed to this email - I
am sure he would love to help out. Keep us in the loop, we would love
to help if you have any graphlab related questions.

Best,

Dr. Danny Bickson
Project Scientist, Machine Learning Dept.
Carnegie Mellon University

[Scott Richardson]

Thanks! i actually found a KernelBP implementation in v2.1's toolkits/graphical_models/deprecated so i can at least take a look at it and see what it does. I'd love to know its current status. 

thanks

Scott