A non-parametric maximum entropy clustering

Hideitsu Hino, Noboru Murata

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

    Abstract

    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 publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    PublisherSpringer Verlag
    Pages113-120
    Number of pages8
    Volume8681 LNCS
    ISBN (Print)9783319111780
    DOIs
    Publication statusPublished - 2014
    Event24th International Conference on Artificial Neural Networks, ICANN 2014 - Hamburg, Germany
    Duration: 2014 Sep 152014 Sep 19

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume8681 LNCS
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other24th International Conference on Artificial Neural Networks, ICANN 2014
    CountryGermany
    CityHamburg
    Period14/9/1514/9/19

    Keywords

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

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

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

    Cite this