Variational Contrastive Log Ratio Upper Bound of Mutual Information for Training Generative Models

Marshal Arijona Sinaga, Machmud Roby Alhamidi, Muhammad Febrian Rachmadi, Wisnu Jatmiko

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Theoretically, a Generative adversarial network minimizes the Jensen-Shannon divergence between real data distribution and generated data distribution. This divergence is another form of mutual information between a mixture distribution and a binary distribution. It implies that we can build a similar generative model by optimizing the mutual information. This research proposes variational contrastive log-ratio upper bound vCLUB mutual information estimation on mixture distribution and the optimization algorithm to train two generative models. We call the models CLUB-sampling generative network (vCLUB-sampling GN) and vCLUB-non sampling generative network (vCLUB-non sampling GN). The results show that vCLUB-sampling outperforms GAN and vCLUB-non sampling GN on the MNIST dataset and has competitive results with GAN on the CIFAR-10 dataset. However, GAN outperforms vCLUB-non sampling GN on both datasets.

Original languageEnglish
Title of host publicationProceedings - IWBIS 2021
Subtitle of host publication6th International Workshop on Big Data and Information Security
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages9-16
Number of pages8
ISBN (Electronic)9781665424516
DOIs
Publication statusPublished - 2021
Event6th International Workshop on Big Data and Information Security, IWBIS 2021 - Virtual, Online, Indonesia
Duration: 23 Oct 202126 Oct 2021

Publication series

NameProceedings - IWBIS 2021: 6th International Workshop on Big Data and Information Security

Conference

Conference6th International Workshop on Big Data and Information Security, IWBIS 2021
Country/TerritoryIndonesia
CityVirtual, Online
Period23/10/2126/10/21

Keywords

  • generative model
  • mutual information
  • neural network
  • variational upper bound minimization

Fingerprint

Dive into the research topics of 'Variational Contrastive Log Ratio Upper Bound of Mutual Information for Training Generative Models'. Together they form a unique fingerprint.

Cite this