Skip to content

Estimation of Mutual Information based on a reverse Jensen inequality approach

License

Notifications You must be signed in to change notification settings

Fritschek/Reverse-Jensen_MI_estimation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Reverese-Jensen_MI_estimation

Estimation of Mutual Information based on a reverse Jensen inequality approach

The RJE (Reverse-Jensen MI Estimator) is based on the recent paper Fritschek et al., "Neural Mutual Information Estimation for Channel Coding: State-of-the-Art Estimators, Analysis, and Performance Comparison", accepted at SPAWC 2020

The code is a shortened version of the colab code by Ben Poole, which shows differences in mutual information estimators, based on their analyses in the paper Poole et al., "On Variational Bounds of Mutual Information", ICML 2019

We removed the interpolation bounds, and unnecessary code due to those bounds, as we want to focus on simpler methods, and 1) added the recent SMILE estimator, based on the paper Song et al., "Understanding the Limitations of Variational Mutual Information Estimators", ICLR 2020, 2) added our RJE Estimator, and 3) added wireless channel models and supporting functions.

About

Estimation of Mutual Information based on a reverse Jensen inequality approach

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published