Connections on non-parametric statistical manifolds by Orlicz space geometry

被引:45
|
作者
Gibilisco, P
Pistone, G
机构
[1] Politecn Torino, Dipartimento Matemat, I-10129 Turin, Italy
[2] Univ Roma Tor Vergata, Ctr Vito Volterra, I-00133 Rome, Italy
关键词
D O I
10.1142/S021902579800017X
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
The non-parametric version of Information Geometry has been developed in recent years. The first basic result was the construction of the manifold structure on M(mu) the maximal statistical models associated to an arbitrary measure mu (see Ref. 48). Using this construction we first show in this paper that the pretangent and the tangent bundles on M(mu) are the natural domains for the mixture connection and for its dual, the exponential connection. Second we show how to define a generalized Amari embedding A(Phi): M(mu) --> S(Phi) from the Exponential Statistical Manifold (ESM) M(mu) to the unit sphere S(Phi) of an arbitrary Orlicz space L(Phi). Finally we show that, in the non-parametric case, the cr-connections del(alpha) (alpha is an element of (-1, 1)) must be defined on a suitable alpha-bundle F(alpha) over M(mu) and that the bundle-connection pair (F(alpha),del(alpha)) is simply (isomorphic to) the pull-back of the Amari embedding A(alpha): M(mu) --> S(2/1-alpha) where the unit sphere S(2/1-alpha)cL(2/1-alpha) is equipped with the natural connection.
引用
收藏
页码:325 / 347
页数:23
相关论文
共 50 条
  • [1] A class of non-parametric statistical manifolds modelled on Sobolev space
    Newton N.J.
    [J]. Information Geometry, 2019, 2 (2) : 283 - 312
  • [2] Riemannian geometry on non-parametric probability space
    Combe-Nencka, H.
    Combe, Ph.
    [J]. NEURAL NETWORK WORLD, 2006, 16 (06) : 459 - 473
  • [3] Statistical non-parametric mapping in sensor space
    Wagner M.
    Tech R.
    Fuchs M.
    Kastner J.
    Gasca F.
    [J]. Biomedical Engineering Letters, 2017, 7 (3) : 193 - 203
  • [4] To be parametric or non-parametric, that is the question Parametric and non-parametric statistical tests
    Van Buren, Eric
    Herring, Amy H.
    [J]. BJOG-AN INTERNATIONAL JOURNAL OF OBSTETRICS AND GYNAECOLOGY, 2020, 127 (05) : 549 - 550
  • [5] STATISTICAL QUESTION Parametric v non-parametric statistical tests
    Sedgwick, Philip
    [J]. BRITISH MEDICAL JOURNAL, 2012, 344
  • [6] Foundations of non-parametric information geometry
    Hasto, P
    Zhang, J
    [J]. JOURNAL OF MATHEMATICAL PSYCHOLOGY, 2005, 49 (01) : 88 - 89
  • [7] Non-parametric statistical fault localization
    Zhang, Zhenyu
    Chan, W. K.
    Tse, T. H.
    Yu, Y. T.
    Hu, Peifeng
    [J]. JOURNAL OF SYSTEMS AND SOFTWARE, 2011, 84 (06) : 885 - 905
  • [8] Statistical inference in the non-parametric case
    Scheffe, H
    [J]. ANNALS OF MATHEMATICAL STATISTICS, 1943, 14 : 305 - 332
  • [9] Non-parametric regression estimation on closed Riemannian manifolds
    Pelletier, B
    [J]. JOURNAL OF NONPARAMETRIC STATISTICS, 2006, 18 (01) : 57 - 67
  • [10] Parametric and non-parametric Statistical Process Control in a dairy industry
    Marra da Silva Ribeiro, Luiz Henrique
    de Araujo, Tatiane Gomes
    Ferreira, Eric Batista
    Zamboni, Jovelino Elias
    [J]. JOURNAL OF CANDIDO TOSTES DAIRY INSTITUTE, 2018, 73 (03): : 112 - 121