Projects authored by zoltan szabo.


Logo JMLR Information Theoretical Estimators 0.62

by szzoli - April 17, 2016, 17:19:00 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 99143 views, 19219 downloads, 3 subscriptions

About: ITE (Information Theoretical Estimators) is capable of estimating many different variants of entropy, mutual information, divergence, association measures, cross quantities and kernels on distributions. Thanks to its highly modular design, ITE supports additionally (i) the combinations of the estimation techniques, (ii) the easy construction and embedding of novel information theoretical estimators, and (iii) their immediate application in information theoretical optimization problems.

Changes:
  • Von Mises expansion based estimators: included for 7 unconditional quantities (Shannon entropy, Shannon mutual information, Kullback-Leibler divergence, Rényi divergence, Tsallis divergence, Pearson Chi^2 divergence, Hellinger distance.

  • Analytical value (for Gaussian random variables) and quick test: added for the Hellinger distance.