Prescription for choosing an interpolating function

DOI10.1063/1.4939710MaRDI QIDQ2795549zbMATH OpenOpenAlexFDO

Authors Tomohisa Takimi

Publication date 21 March 2016

Published in Journal of Mathematical Physics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1410.1484

zbMATH Keywords

correlation coefficientdeviation of actual function to its interpolantgood interpolating functionreference quantities

Mathematics Subject Classification ID

Interpolation in approximation theory (41A05) Padé approximation (41A21) Alternative quantum mechanics (including hidden variables, etc.) (81Q65)

Abstract: Interpolating functional method is a powerful tool for studying the behavior of a quantity in the intermediate region of the parameter space of interest by using its perturbative expansions at both ends. Recently several interpolating functional methods have been proposed, in addition to the well-known Pade approximant, namely the "Fractional Power of Polynomial" (FPP) and the "Fractional Power of Rational functions" (FPR) methods. Since combinations of these methods also give interpolating functions, we may end up with multitudes of the possible approaches. So a criterion for choosing an appropriate interpolating function is very much needed. In this paper, we propose reference quantities which can be used for choosing a good interpolating function. In order to validate the prescription based on these quantities, we study the degree of correlation between "the reference quantities" and the "actual degree of deviation between the interpolating function and the true function" in examples where the true functions are known.

Review text

The author proposes several quantities as the reference quantities for selecting a good interpolating function. Correlation between the set of these quantities and the actual deviation between the true function \(F(g)\) and its interpolating function \(G(g)\), where \(g\) is a parameter of interest is important to check the quality of references. Considering examples where the actual function is known the author examines the reference quantities by computing the correlation coefficients. Stating that a criterion suggested in [\textit{M. Honda}, J. High Energy Phys. 2014, No. 12, Paper No. 019, 43 p. (2014; Zbl 1338.81288)] was insufficient the authors observe that as a first step they worked by using several examples only and that further studies were needed to establish a very general criterion.

Recommendations

• scientific article; zbMATH DE number 733429
• The best parameterization for parametric interpolation
• An alternative procedure for selecting a good value for the parameter \(c\) in RBF-interpolation
• An algorithm for selecting a good value for the parameter \(c\) in radial basis function interpolation
• scientific article; zbMATH DE number 6613320

Cites work

• Critical properties of ^4-theories
• Generalized bootstrap equations for \( \mathcal{N}=4 \) SCFT
• High and low dimensions in the black hole negative mode
• Lectures on non-perturbative effects in large N gauge theories, matrix models and strings
• On perturbation theory improved by strong coupling expansion
• S-duality improved perturbation theory in compactified type I/heterotic string theory
• S-duality improved superstring perturbation theory

Cited in

(2)

• On perturbation theory improved by strong coupling expansion
• Interpolating function and Stokes phenomena

This page was built for publication: Prescription for choosing an interpolating function

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2795549)