Infinite-dimensional statistical manifolds based on a balanced chart

From MaRDI portal
Publication:265266

DOI10.3150/14-BEJ673zbMATH Open1388.60043arXiv1308.3602MaRDI QIDQ265266FDOQ265266


Authors: Nigel J. Newton Edit this on Wikidata


Publication date: 1 April 2016

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We develop a family of infinite-dimensional Banach manifolds of measures on an abstract measurable space, employing charts that are "balanced" between the density and log-density functions. The manifolds, (ildeMlambda,lambdain[2,infty)), retain many of the features of finite-dimensional information geometry; in particular, the alpha-divergences are of class Clceillambdaceil1, enabling the definition of the Fisher metric and alpha-derivatives of particular classes of vector fields. Manifolds of probability measures, (Mlambda,lambdain[2,infty)), based on centred versions of the charts are shown to be Clceillambdaceil1-embedded submanifolds of the ildeMlambda. The Fisher metric is a pseudo-Riemannian metric on ildeMlambda. However, when restricted to finite-dimensional embedded submanifolds it becomes a Riemannian metric, allowing the full development of the geometry of alpha-covariant derivatives. ildeMlambda and Mlambda provide natural settings for the study and comparison of approximations to posterior distributions in problems of Bayesian estimation.


Full work available at URL: https://arxiv.org/abs/1308.3602




Recommendations




Cites Work


Cited In (9)





This page was built for publication: Infinite-dimensional statistical manifolds based on a balanced chart

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q265266)