Skip to main navigation Skip to search Skip to main content

Convergence of Binarized Context-tree Weighting for Estimating Distributions of Stationary Sources

Badri N. Vellambi, Marcus Hutter

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

    2 Citations (Scopus)

    Abstract

    This work investigates the convergence rate of learning the stationary distribution of finite-alphabet stationary ergodic sources using a binarized context-tree weighting approach. The binarized context-tree weighting (overline mathbf Cmathbf Tmathbf W) algorithm estimates the stationary distribution of a symbol as a product of conditional distributions of each component bit, which are determined in a sequential manner using the well known binary context-tree weighting method.

    Original languageEnglish
    Title of host publication2018 IEEE International Symposium on Information Theory, ISIT 2018
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages731-735
    Number of pages5
    ISBN (Print)9781538647806
    DOIs
    Publication statusPublished - 15 Aug 2018
    Event2018 IEEE International Symposium on Information Theory, ISIT 2018 - Vail, United States
    Duration: 17 Jun 201822 Jun 2018

    Publication series

    NameIEEE International Symposium on Information Theory - Proceedings
    Volume2018-June
    ISSN (Print)2157-8095

    Conference

    Conference2018 IEEE International Symposium on Information Theory, ISIT 2018
    Country/TerritoryUnited States
    CityVail
    Period17/06/1822/06/18

    Fingerprint

    Dive into the research topics of 'Convergence of Binarized Context-tree Weighting for Estimating Distributions of Stationary Sources'. Together they form a unique fingerprint.

    Cite this