Strong entropy concentration, game theory and algorithmic randomness

P.D. Grünwald

    Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

    5 Citaten (Scopus)
    2 Downloads (Pure)

    Samenvatting

    We give a characterization of Maximum Entropy/Minimum Relative Entropy inference by providing two ‘strong entropy concentration’ theorems. These theorems unify and generalize Jaynes’ ‘concentration phenomenon’ and Van Campenhout and Cover’s ‘conditional limit theorem’. The theorems characterize exactly in what sense a ‘prior’ distribution Q conditioned on a given constraint and the distribution P minimizing D(P//Q) over all P satisfyingthe constraint are ‘close’ to each other. We show how our theorems are related to ‘universal models’ for exponential families, thereby establishinga link with Rissanen’s MDL/stochastic complexity. We then apply our theorems to establish the relationship (A) between entropy concentration and a game-theoretic characterization of Maximum Entropy Inference due to Topsøe and others; (B) between maximum entropy distributions and sequences that are random (in the sense of Martin-Löf/Kolmogorov) with respect to the given constraint. These two applications have strong implications for the use of Maximum Entropy distributions in sequential prediction tasks, both for the logarithmic loss and for general loss functions. We identify circumstances under which Maximum Entropy predictions are almost optimal.
    Originele taal-2Engels
    TitelComputational learning theory : proceedings 14th annual conference on computational learning theory, COLT 2001, and 5th European conference on computational learning theory, EuroCOLT 2001, Amsterdam, The Netherlands, july 16-19, 2001
    RedacteurenD.P. Helmbold, B. Williamson
    Plaats van productieBerlin
    UitgeverijSpringer
    Pagina's320-336
    ISBN van geprinte versie3-540-42343-5
    DOI's
    StatusGepubliceerd - 2001

    Publicatie series

    NaamLecture Notes in Computer Science
    Volume2111
    ISSN van geprinte versie0302-9743

    Vingerafdruk Duik in de onderzoeksthema's van 'Strong entropy concentration, game theory and algorithmic randomness'. Samen vormen ze een unieke vingerafdruk.

    Citeer dit