A Lecture About the Use of Orlicz Spaces in Information Geometry

Giovanni Pistone

DOI: https://doi.org/10.1007/978-3-030-77957-3_10

2021-01-01

Abstract:This chapter is a revised version of a tutorial lecture that I presented at the École de Physique des Houches on July 26–31 2020. Topics include: Non-parametric Information Geometry, the Statistical bundle, exponential Orlicz spaces, and Gaussian Orlicz-Sobolev spaces.

What problem does this paper attempt to address?

Translations in the exponential Orlicz space with Gaussian weight

Giovanni Pistone

DOI: https://doi.org/10.48550/arXiv.1708.09555

2017-08-31

Statistics Theory

Abstract:We study the continuity of space translations on non-parametric exponential families based on the exponential Orlicz space with Gaussian reference density.
Information Geometry of the Probability Simplex: A Short Course

Giovanni Pistone

DOI: https://doi.org/10.33581/1561-4085-2020-23-2-221-242

2020-07-09

Nonlinear Phenomena in Complex Systems

Abstract:This set of notes is intended for a short course aiming to provide an (almost) self-contained and (almost) elementary introduction to the topic of Information Geometry (IG) of the probability simplex. Such a course can be considered an introduction to the original monograph by Amari and Nagaoka [1], and to the recent monographs by Amari [2] and by Ay, Jost, Lê, and Schwachhöfer [3]. The focus is on a non-parametric approach, that is, I consider the geometry of the full probability simplex and compare the IG formalism with what is classically done in Statistical Physics.
Information Geometry of the Gaussian Space

Giovanni Pistone

DOI: https://doi.org/10.48550/arXiv.1803.08135

2018-03-20

Probability

Abstract:We discuss the Pistone-Sempi exponential manifold on the finite-dimensional Gaussian space. We consider the role of the entropy, the continuity of translations, Poincar\'e-type inequalities, the generalized differentiability of probability densities of the Gaussian space.
Affine statistical bundle modeled on a Gaussian Orlicz-Sobolev space

Giovanni Pistone

DOI: https://doi.org/10.48550/arXiv.2210.07641

2022-10-14

Abstract:The dually flat structure of statistical manifolds can be derived in a non-parametric way from a particular case of affine space defined on a qualified set of probability measures. The statistically natural displacement mapping of the affine space depends on the notion of Fisher's score. The model space must be carefully defined if the state space is not finite. Among various options, we discuss how to use Orlicz-Sobolev spaces with Gaussian weight. Such a fully non-parametric set-up provides tools to discuss intrinsically infinite-dimensional evolution problems.

Statistics Theory,Differential Geometry,Functional Analysis
Examples of Application of Nonparametric Information Geometry to Statistical Physics

Giovanni Pistone

DOI: https://doi.org/10.48550/arXiv.1308.5312

2013-08-24

Statistics Theory

Abstract:We review a nonparametric version of Amari's Information Geometry in which the set of positive probability densities on a given sample space is endowed with an atlas of charts to form a differentiable manifold modeled on Orlicz Banach spaces. This nonparametric setting is used to discuss the setting of typical problems in Machine Learning and Statistical Physics, such as relaxed optimization, Kullback-Leibler divergence, Boltzmann entropy, Boltzmann equation
The exponential Orlicz space in quantum information geometry

Anna Jenčová

DOI: https://doi.org/10.1007/s41884-023-00097-x

2023-01-24

Information Geometry

Abstract:We review the construction of a quantum version of the exponential statistical manifold over the set of all faithful normal positive functionals on a von Neumann algebra. The construction is based on the relative entropy approach to state perturbation. We construct a quantum version of the exponential Orlicz space and discuss the properties of this space and its dual with respect to Kosaki Lpdocumentclass[12pt]{minimal}usepackage{amsmath}usepackage{wasysym}usepackage{amsfonts}usepackage{amssymb}usepackage{amsbsy}usepackage{mathrsfs}usepackage{upgreek}setlength{oddsidemargin}{-69pt}egin{document}$$L_p$$end{document}-spaces. We show that the constructed manifold admits a canonical divergence satisfying a Pythagorean relation. We also prove that the manifold structure is invariant under sufficient channels.
Optimization via Information Geometry

Luigi Malagò,Giovanni Pistone

DOI: https://doi.org/10.48550/arXiv.1403.3762

2014-03-15

Statistics Theory

Abstract:Information Geometry has been used to inspire efficient algorithms for stochastic optimization, both in the combinatorial and the continuous case. We give an overview of the authors' research program and some specific contributions to the underlying theory.
Lectures on families of Dirac operators and applications

Francesco Lin

DOI: https://doi.org/10.48550/arXiv.2209.01467

2022-09-04

Abstract:These are the notes for a minicourse taught at the 2022 ICTP summer school `Frontiers in Geometry and Topology'. The goal is to introduce families of Dirac operators and how they can be used to study interactions between geometry and topology. In particular, we discuss the index theorem for families of twisted Dirac operators, and its applications to metrics of positive scalar curvature and the three-dimensional Weinstein conjecture.

Differential Geometry,Geometric Topology,Symplectic Geometry
Orlicz spaces and the large scale geometry of Heintze groups

Matias Carrasco Piaggio

DOI: https://doi.org/10.48550/arXiv.1411.6381

2014-11-24

Abstract:We consider an Orlicz space based cohomology for metric (measured) spaces with bounded geometry. We prove the quasi-isometry invariance for a general Young function. In the hyperbolic case, we prove that the degree one cohomology can be identified with an Orlicz-Besov function space on the boundary at infinity. We give some applications to the large scale geometry of homogeneous spaces with negative curvature (Heintze groups). As our main result, we prove that if the Heintze group is not of Carnot type, any self quasi-isometry fixes a distinguished point on the boundary and preserves a certain foliation on the complement of that point.

Metric Geometry
Dually affine Information Geometry modeled on a Banach space

Goffredo Chirco,Giovanni Pistone

2024-05-13

Abstract:In this chapter, we study Information Geometry from a particular non-parametric or functional point of view. The basic model is a probabilities subset usually specified by regularity conditions. For example, probability measures mutually absolutely continuous or probability densities with a given degree of smoothness. We construct a manifold structure by giving an atlas of charts as mappings from probabilities to a Banach space. The charts we use are quite peculiar in that we consider only instances where the transition mappings are affine. We chose a particular expression of the tangent and cotangent bundles in this affine setting.

Statistics Theory
Information Geometry for the Working Information Theorist

Kumar Vijay Mishra,M. Ashok Kumar,Ting-Kam Leonard Wong

2023-10-06

Abstract:Information geometry is a study of statistical manifolds, that is, spaces of probability distributions from a geometric perspective. Its classical information-theoretic applications relate to statistical concepts such as Fisher information, sufficient statistics, and efficient estimators. Today, information geometry has emerged as an interdisciplinary field that finds applications in diverse areas such as radar sensing, array signal processing, quantum physics, deep learning, and optimal transport. This article presents an overview of essential information geometry to initiate an information theorist, who may be unfamiliar with this exciting area of research. We explain the concepts of divergences on statistical manifolds, generalized notions of distances, orthogonality, and geodesics, thereby paving the way for concrete applications and novel theoretical investigations. We also highlight some recent information-geometric developments, which are of interest to the broader information theory community.

Information Theory,Machine Learning,Signal Processing,Differential Geometry
Lectures on Linear Series

Robert Lazarsfeld

DOI: https://doi.org/10.48550/arXiv.alg-geom/9408011

1994-09-01

Abstract:The past decade has witnessed two important new developments in the study of linear series on algebraic varieties. First, vector bundles have emerged as powerful tools for analyzing linear series on curves and surfaces. More recently, the viewpoints and techniques of higher dimensional geometry have started to play a central role. The purpose of these lectures is to provide a down-to-earth introduction to some of these new ideas and methods. The present notes, written with the assistance of G. Fernandez del Busto, are a considerably expanded version of a course given at the 1993 Regional Geometry Institute on Higher Dimensional Complex Geometry at Park City, Utah. While I hope they may be of interest to any geometers wishing to learn about recent work on linear series, I have particularly attempted to aim the discusion at a novice audience. My hope is that with a little faith and effort, these notes should be accessible to anyone having finished the standard introductory texts in algebraic geometry. The underlying theme of the lectures is the search for higher dimensional generalizations of the most basic facts about linear series on curves. But in an attempt to keep things elementary we work more or less exclusively on surfaces. Thus we end up discussing one central result -- viz. Reider's theorem -- from many points of view. Sticking to the case of surfaces allows one to eliminate many technical complexities, and I hope that parts of these lectures might therefore serve as a useful first introduction to the powerful cohomological techniques (involving $Q$-divisors) of contemporary

Algebraic Geometry
Algebra and geometry of link homology

Eugene Gorsky,Oscar Kivinen,José Simental

DOI: https://doi.org/10.48550/arXiv.2108.10356

2021-08-24

Abstract:These notes cover the lectures of the first named author at 2021 IHES Summer School on "Enumerative Geometry, Physics and Representation Theory" with additional details and references. They cover the definition of Khovanov-Rozansky triply graded homology, its basic properties and recent advances, as well as three algebro-geometric models for link homology: braid varieties, Hilbert schemes of singular curves and affine Springer fibers, and Hilbert schemes of points on the plane.

Algebraic Geometry,Combinatorics,Geometric Topology,Representation Theory
GGI Lectures on Entropy, Operator Algebras and Black Holes

Stefan Hollands

DOI: https://doi.org/10.48550/arXiv.2209.05132

2023-03-08

Abstract:These are slightly extended notes for my lectures at the workshop ``Reconstructing the Gravitational Hologram'' held at Galileo-Galilei Institute, Florence in June 2022, v2 changes to sec. 9 on generalized second law

High Energy Physics - Theory
Geometric Modeling in Probability and Statistics

Ovidiu Calin,Constantin Udrişte

DOI: https://doi.org/10.1007/978-3-319-07779-6

2014-01-01

Abstract:This book covers topics of Informational Geometry, a field which deals with the differential geometric study of the manifold probability density functions. This is a field that is increasingly attracting the interest of researchers from many different areas of science, including mathematics, statistics, geometry, computer science, signal processing, physics and neuroscience. It is the authors’ hope that the present book will be a valuable reference for researchers and graduate students in one of the aforementioned fields. This textbook is a unified presentation of differential geometry and probability theory, and constitutes a text for a course directed at graduate or advanced undergraduate students interested in applications of differential geometry in probability and statistics. The book contains over 100 proposed exercises meant to help students deepen their understanding, and it is accompanied by software that is able to provide numerical computations of several information geometric objects. The reader will understand a flourishing field of mathematics in which very few books have been written so far.
Adic spaces

Katharina Hübner

2024-05-10

Abstract:These lecture notes are based on the second course in a series of lectures at the Spring school "Non-archimedean geometry and Eigenvarieties" in March 2023 in Heidelberg. The objective of the first three courses was to give an introduction to the theory of adic spaces. Building up on the theory of Huber pairs presented in John Bergdall's lecture we explain the construction of adic spaces. We study some important classes of adic spaces such as rigid analytic spaces and formal schemes and show the connections between them. In the course of the lecture we will illustrate the respective concepts with the fundamental examples of the open and closed disc and the affine line.

Number Theory
Understanding Higher-Order Interactions in Information Space

Herbert Edelsbrunner,Katharina Ölsböck,Hubert Wagner

DOI: https://doi.org/10.3390/e26080637

2024-07-27

Abstract:Methods used in topological data analysis naturally capture higher-order interactions in point cloud data embedded in a metric space. This methodology was recently extended to data living in an information space, by which we mean a space measured with an information theoretical distance. One such setting is a finite collection of discrete probability distributions embedded in the probability simplex measured with the relative entropy (Kullback-Leibler divergence). More generally, one can work with a Bregman divergence parameterized by a different notion of entropy. While theoretical algorithms exist for this setup, there is a paucity of implementations for exploring and comparing geometric-topological properties of various information spaces. The interest of this work is therefore twofold. First, we propose the first robust algorithms and software for geometric and topological data analysis in information space. Perhaps surprisingly, despite working with Bregman divergences, our design reuses robust libraries for the Euclidean case. Second, using the new software, we take the first steps towards understanding the geometric-topological structure of these spaces. In particular, we compare them with the more familiar spaces equipped with the Euclidean and Fisher metrics.
Lectures on Poisson algebras

Volodya Roubtsov,Radek Suchánek

2023-05-05

Abstract:Poisson algebras have become an essential topic in mathematics with a rich structure and wide applicability. Despite numerous resources available on Poisson structures, the algebraic side of the story remains relatively less explored. This paper presents an extended version of a mini-course given during the virtual Winter School and Workshop "Wisla 20-21." The course covers various aspects of Poisson algebras, focusing on both algebraic and geometric perspectives. While most of the lectures cover established results, the final chapter includes recent classification results on "low rank" double quadratic Poisson brackets.

Rings and Algebras,Commutative Algebra
Two Lectures on Iterative Coding and Statistical Mechanics

Andrea Montanari

DOI: https://doi.org/10.48550/arXiv.cond-mat/0512296

2005-12-14

Abstract:These are the notes for two lectures delivered at the Les Houches summer school Mathematical Statistical Mechanics, held in July 2005. I review some basic notions on sparse graph error correcting codes with emphasis on `modern' aspects, such as, iterative belief propagation decoding. Relations with statistical mechanics, inference and random combinatorial optimization are stressed, as well as some general mathematical ideas and open problems.

Statistical Mechanics,Disordered Systems and Neural Networks
Distortion of Hausdorff measures under Orlicz--Sobolev maps

Andrea Cianchi,Mikhail V. Korobkov,Jan Kristensen

DOI: https://doi.org/10.48550/arxiv.2208.08152

2022-01-01

Abstract: A comprehensive theory of the effect of Orlicz-Sobolev maps, between Euclidean spaces, on subsets with zero or finite Hausdorff measure is offered. Arbitrary Orlicz-Sobolev spaces embedded into the space of continuous function and Hausdorff measures built upon general gauge functions are included in our discussion. An explicit formula for the distortion of the relevant gauge function under the action of these maps is exhibited in terms of the Young function defining the Orlicz-Sobolev space. New phenomena and features, related to the flexibility in the definition of the degree of integrability of weak derivatives of maps and in the notion of measure of sets, are detected. Classical results, dealing with standard Sobolev spaces and Hausdorff measures, are recovered, and their optimality is shown to hold in a refined stronger sense. Special instances available in the literature, concerning Young functions and gauge functions of non-power type, are also reproduced and, when not sharp, improved.

A Lecture About the Use of Orlicz Spaces in Information Geometry

Translations in the exponential Orlicz space with Gaussian weight

Information Geometry of the Probability Simplex: A Short Course

Information Geometry of the Gaussian Space

Affine statistical bundle modeled on a Gaussian Orlicz-Sobolev space

Examples of Application of Nonparametric Information Geometry to Statistical Physics

The exponential Orlicz space in quantum information geometry

Optimization via Information Geometry

Lectures on families of Dirac operators and applications

Orlicz spaces and the large scale geometry of Heintze groups

Dually affine Information Geometry modeled on a Banach space

Information Geometry for the Working Information Theorist

Lectures on Linear Series

Algebra and geometry of link homology

GGI Lectures on Entropy, Operator Algebras and Black Holes

Geometric Modeling in Probability and Statistics

Adic spaces

Understanding Higher-Order Interactions in Information Space

Lectures on Poisson algebras

Two Lectures on Iterative Coding and Statistical Mechanics

Distortion of Hausdorff measures under Orlicz--Sobolev maps