A non-parametric maximum entropy clustering

Hideitsu Hino, Noboru Murata

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Clustering is a fundamental tool for exploratory data analysis. Information theoretic clustering is based on the optimization of information theoretic quantities such as entropy and mutual information. Recently, since these quantities can be estimated in non-parametric manner, non-parametric information theoretic clustering gains much attention. Assuming the dataset is sampled from a certain cluster, and assigning different sampling weights depending on the clusters, the cluster conditional information theoretic quantities are estimated. In this paper, a simple clustering algorithm is proposed based on the principle of maximum entropy. The algorithm is experimentally shown to be comparable to or outperform conventional non-parametric clustering methods.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning, ICANN 2014 - 24th International Conference on Artificial Neural Networks, Proceedings
PublisherSpringer Verlag
Number of pages8
ISBN (Print)9783319111780
Publication statusPublished - 2014
Event24th International Conference on Artificial Neural Networks, ICANN 2014 - Hamburg, Germany
Duration: 2014 Sept 152014 Sept 19

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8681 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other24th International Conference on Artificial Neural Networks, ICANN 2014


  • Information Theoretic Clustering
  • Likelihood and Entropy estimator
  • Non-parametric

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'A non-parametric maximum entropy clustering'. Together they form a unique fingerprint.

Cite this