Skip to content
The demo for "Convolutional Poisson Gamma Belief Network" published in ICML2019
Jupyter Notebook Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

==Convolutional Poisson Gamma Belief Network==

This is code for the paper "Convolutional Poisson Gamma Belief Network" published in ICML2019.

Created by Chaojie Wang , Bo Chen , Sucheng Xiao at Xidian University and Mingyuan Zhou at University of Texas at Austin.


Tensorflow >= 1.0

PyCUDA >= 0.8

PyCUDA can be download from following address

==Data Source==

All data source files can be found in following addresses and have been included in our repository.


  • CPFA_Mnist_Demo folder contains 4 different training algorithms for CPFA, including Toeplitz, Element, Element-Parallel and SGMCMC mehthods.

  • CPGBN_Text_Demo folder contains Datasets and experiment code to reproduce the results in our paper.

  • CPGBN_Derivation_Draft file provides a detailed derivation for the CPGBN.


If you find that the algorithms in this repository are useful for your research, please refer to the following article:

title={{C}onvolutional {P}oisson {G}amma {B}elief {N}etwork},
author={Chaojie Wang and Bo Chen and Sucheng Xiao and Mingyuan Zhou}, booktitle={ICML}, year={2019}}


Contact Bo Chen or Chaojie Wang

Copyright (c), 2018, Chaojie Wang

You can’t perform that action at this time.