The Tempered Hilbert Simplex Distance and Its Application To Non-linear Embeddings of TEMs

• URL: http://arxiv.org/abs/2311.13459v1
• Date: Wed, 22 Nov 2023 15:24:29 GMT
• Title: The Tempered Hilbert Simplex Distance and Its Application To Non-linear Embeddings of TEMs
• Authors: Ehsan Amid, Frank Nielsen, Richard Nock, Manfred K. Warmuth
• Abstract summary: We introduce three different parameterizations of finite discrete TEMs via Legendre functions of the negative tempered entropy function. Similar to the Hilbert geometry, the tempered Hilbert distance is characterized as a $t$-symmetrization of the oriented tempered Funk distance.
• Score: 36.135201624191026
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Tempered Exponential Measures (TEMs) are a parametric generalization of the exponential family of distributions maximizing the tempered entropy function among positive measures subject to a probability normalization of their power densities. Calculus on TEMs relies on a deformed algebra of arithmetic operators induced by the deformed logarithms used to define the tempered entropy. In this work, we introduce three different parameterizations of finite discrete TEMs via Legendre functions of the negative tempered entropy function. In particular, we establish an isometry between such parameterizations in terms of a generalization of the Hilbert log cross-ratio simplex distance to a tempered Hilbert co-simplex distance. Similar to the Hilbert geometry, the tempered Hilbert distance is characterized as a $t$-symmetrization of the oriented tempered Funk distance. We motivate our construction by introducing the notion of $t$-lengths of smooth curves in a tautological Finsler manifold. We then demonstrate the properties of our generalized structure in different settings and numerically examine the quality of its differentiable approximations for optimization in machine learning settings.

Related papers

• Symmetry & Critical Points for Symmetric Tensor Decomposition Problems [6.650108973968032]
  We consider the non optimization problem associated with the decomposition of a real symmetric tensor into a sum of rank one terms. Use is made of a rich symmetry structure to construct infinite families of critical points represented by Puiseux series in the problem dimension. Adesirable phenomenon, occurring for all critical points considered, concerns the number of negative Hessian eigenvalues increasing with the value of the objective function.
  arXiv  Detail & Related papers  (2023-06-13T16:25:30Z)
• Kernel-based off-policy estimation without overlap: Instance optimality beyond semiparametric efficiency [53.90687548731265]
  We study optimal procedures for estimating a linear functional based on observational data. For any convex and symmetric function class $mathcalF$, we derive a non-asymptotic local minimax bound on the mean-squared error.
  arXiv  Detail & Related papers  (2023-01-16T02:57:37Z)
• Coherence generation, symmetry algebras and Hilbert space fragmentation [0.0]
  We show a simple connection between classification of physical systems and their coherence generation properties, quantified by the coherence generating power (CGP) We numerically simulate paradigmatic models with both ordinary symmetries and Hilbert space fragmentation, comparing the behavior of the CGP in each case with the system dimension.
  arXiv  Detail & Related papers  (2022-12-29T18:31:16Z)
• Kullback-Leibler and Renyi divergences in reproducing kernel Hilbert space and Gaussian process settings [0.0]
  We present formulations for regularized Kullback-Leibler and R'enyi divergences via the Alpha Log-Determinant (Log-Det) divergences. For characteristic kernels, the first setting leads to divergences between arbitrary Borel probability measures on a complete, separable metric space. We show that the Alpha Log-Det divergences are continuous in the Hilbert-Schmidt norm, which enables us to apply laws of large numbers for Hilbert space-valued random variables.
  arXiv  Detail & Related papers  (2022-07-18T06:40:46Z)
• Rational Approximations of Quasi-Periodic Problems via Projected Green's Functions [0.0]
  We introduce the projected Green's function technique to study quasi-periodic systems. The technique is flexible and can be used to extract both analytic and numerical results.
  arXiv  Detail & Related papers  (2021-09-28T18:00:00Z)
• Complexity-like properties and parameter asymptotics of $\mathfrak{L}_{q}$-norms of Laguerre and Gegenbauer polynomials [0.0]
  Main monotonic statistical complexity-like measures of the Rakhmanov's probability density associated to the hypergeometrics (HOP) in a real continuous variable. The degree and parameters of these two-fold spreading measures are shown for the parameter-dependent families of HOPs of Laguerre and Gegenbauer types.
  arXiv  Detail & Related papers  (2021-08-16T16:49:49Z)
• q-Paths: Generalizing the Geometric Annealing Path using Power Means [51.73925445218366]
  We introduce $q$-paths, a family of paths which includes the geometric and arithmetic mixtures as special cases. We show that small deviations away from the geometric path yield empirical gains for Bayesian inference.
  arXiv  Detail & Related papers  (2021-07-01T21:09:06Z)
• Entanglement Entropy of Non-Hermitian Free Fermions [59.54862183456067]
  We study the entanglement properties of non-Hermitian free fermionic models with translation symmetry. Our results show that the entanglement entropy has a logarithmic correction to the area law in both one-dimensional and two-dimensional systems.
  arXiv  Detail & Related papers  (2021-05-20T14:46:09Z)
• Exact thermal properties of free-fermionic spin chains [68.8204255655161]
  We focus on spin chain models that admit a description in terms of free fermions. Errors stemming from the ubiquitous approximation are identified in the neighborhood of the critical point at low temperatures.
  arXiv  Detail & Related papers  (2021-03-30T13:15:44Z)
• Optimal oracle inequalities for solving projected fixed-point equations [53.31620399640334]
  We study methods that use a collection of random observations to compute approximate solutions by searching over a known low-dimensional subspace of the Hilbert space. We show how our results precisely characterize the error of a class of temporal difference learning methods for the policy evaluation problem with linear function approximation.
  arXiv  Detail & Related papers  (2020-12-09T20:19:32Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.