The Taylor series of the Gaussian kernel

L. Escauriaza

DOI: https://doi.org/10.48550/arXiv.math/0606035

2006-10-04

Abstract:We describe a formula for the Taylor series expansion of the Gaussian kernel around the origin of $\Rn\times\R$.

Analysis of PDEs

What problem does this paper attempt to address?

A probabilistic Taylor expansion with Gaussian processes

Toni Karvonen,Jon Cockayne,Filip Tronarp,Simo Särkkä

2023-08-28

Abstract:We study a class of Gaussian processes for which the posterior mean, for a particular choice of data, replicates a truncated Taylor expansion of any order. The data consist of derivative evaluations at the expansion point and the prior covariance kernel belongs to the class of Taylor kernels, which can be written in a certain power series form. We discuss and prove some results on maximum likelihood estimation of parameters of Taylor kernels. The proposed framework is a special case of Gaussian process regression based on data that is orthogonal in the reproducing kernel Hilbert space of the covariance kernel.

Machine Learning,Numerical Analysis,Methodology
On the Gaussian Kinematic Formula of R. Adler and J. Taylor

Joseph H.G. Fu

2024-09-19

Abstract:We apply methods of algebraic integral geometry to prove a special case of the Gaussian kinematic formula of Adler-Taylor. The idea, suggested already by Adler and Taylor, is to view the GKF as the limit of spherical kinematic formulas for spheres of large dimension $N$ and curvature $\frac 1 N$.

Differential Geometry
Exponential Taylor Series

André Kowacs

2024-08-23

Abstract:This paper derives a way to express differentiable complex-valued functions as the sum of powers of $(1-e^{\lambda x})$, where $\lambda\in\mathbb{R}$, with an explicit formula for the remainder. This formulation is then used to associate an infinite series to $C^\infty$ functions, which is shown to recover the original function under suitable conditions on the remainder. These results are also used to calculate some infinite series involving Stirling Numbers, as well as providing a few examples.

Classical Analysis and ODEs
Taylor expansion for an operator function

Ioan Sturzu

DOI: https://doi.org/10.48550/arXiv.math-ph/0408059

2004-08-31

Abstract:A simple version for the extension of the Taylor theorem to the operator functions was found. The expansion was done with respect to a value given by a diagonal matrix for the non-commutative case, and the coefficients are given both by recurrence relations and Cauchy integrals.

Mathematical Physics
On the Asymptotic Expansion of Bergman Kernel

XZ Dai,KF Liu,XN Ma

DOI: https://doi.org/10.1016/j.crma.2004.05.011

2004-01-01

Comptes Rendus Mathématique

Abstract:We study the asymptotic of the Bergman kernel of the spin$^c$ Dirac operator on high tensor powers of a line bundle.
Taylor's Law from Gaussian diffusions

Iddo Isaac Eliazar

DOI: https://doi.org/10.1088/1751-8121/ad952d

2024-11-28

Journal of Physics A Mathematical and Theoretical

Abstract:Taylor's Law, also known as fluctuation scaling, manifests a power relation between the means and the variances of statistical distributions. The class of Gaussian-selfsimilar stochastic motions offers a plethora of macroscopic diffusion models, regular and anomalous alike. This class includes Brownian motion, scaled Brownian motion, fractional Brownian motion, and more. Within this class, power Brownian motion is the sub-class of motions that are also Markovian. Considering conditional distributions of motion positions, this paper establishes that: the Gaussian-selfsimilar class universally generates Taylor's Law, doing so with both positive and negative Taylor exponents. The paper also unveils a profound interplay between power Brownian motion, and the universal generation of Taylor's Law from the Gaussian-selfsimilar class.

physics, multidisciplinary, mathematical
Taylor series for the Askey-Wilson operator and classical summation formulas

José Manuel Marco,Javier Parcet

DOI: https://doi.org/10.48550/arXiv.math/0312248

2005-08-26

Abstract:An analogue of Taylor's formula, which arises by substituting the classical derivative by a divided difference operator of Askey-Wilson type, is developed here. We study the convergence of the associated Taylor series. Our results complement a recent work by Ismail and Stanton. Quite surprisingly, in some cases the Taylor polynomials converge to a function which differs from the original one. We provide explicit expressions for the integral remainder. As application, we obtain some summation formulas for basic hypergeometric series. As far as we know, one of them is new. We conclude by studying the different forms of the binomial theorem in this context.

Classical Analysis and ODEs
Taylor's Formula for Generalized Weighted Fractional Derivatives with Nonsingular Kernels

Houssine Zine,El Mehdi Lotfi,Delfim F. M. Torres,Noura Yousfi

DOI: https://doi.org/10.3390/axioms11050231

2022-05-16

Axioms

Abstract:We prove a new Taylor's theorem for generalized weighted fractional calculus with nonsingular kernels. The proof is based on the establishment of new relations for nth-weighted generalized fractional integrals and derivatives. As an application, new mean value theorems for generalized weighted fractional operators are obtained. Direct corollaries allow one to obtain the recent Taylor's and mean value theorems for Caputo–Fabrizio, Atangana–Baleanu–Caputo (ABC) and weighted ABC derivatives.
Generalized Taylor's formula for power fractional derivatives

Hanaa Zitane,Delfim F. M. Torres

DOI: https://doi.org/10.1007/s40590-023-00540-0

2023-09-08

Abstract:We establish a new generalized Taylor's formula for power fractional derivatives with nonsingular and nonlocal kernels, which includes many known Taylor's formulas in the literature. Moreover, as a consequence, we obtain a general version of the classical mean value theorem. We apply our main result to approximate functions in Taylor's expansions at a given point. The explicit interpolation error is also obtained. The new results are illustrated through examples and numerical simulations.

Spectral Theory
On Taylor series and Kapteyn series of the first and second type

Diego Dominici

DOI: https://doi.org/10.48550/arXiv.1012.3700

2010-12-17

Abstract:We study the relation between the coefficients of Taylor series and Kapteyn series representing the same function. We compute explicit formulas for expressing one in terms of the other and give examples to illustrate our method.

Classical Analysis and ODEs
On Taylor Series Expansion for Statistical Moments of Functions of Correlated Random Variables

Lukáš Novák,Drahomír Novák

DOI: https://doi.org/10.3390/sym12081379

2020-08-18

Symmetry

Abstract:The paper is focused on Taylor series expansion for statistical analysis of functions of random variables with special attention to correlated input random variables. It is shown that the standard approach leads to significant deviations in estimated variance of non-linear functions. Moreover, input random variables are often correlated in industrial applications; thus, it is crucial to obtain accurate estimations of partial derivatives by a numerical differencing scheme. Therefore, a novel methodology for construction of Taylor series expansion of increasing complexity of differencing schemes is proposed and applied on several analytical examples. The methodology is adapted for engineering applications by proposed asymmetric difference quotients in combination with a specific step-size parameter. It is shown that proposed differencing schemes are suitable for functions of correlated random variables. Finally, the accuracy, efficiency, and limitations of the proposed methodology are discussed.
Images of Gaussian and other stochastic processes under closed, densely-defined, unbounded linear operators

Tadashi Matsumoto,T. J. Sullivan

2024-01-12

Abstract:Gaussian processes (GPs) are widely-used tools in spatial statistics and machine learning and the formulae for the mean function and covariance kernel of a GP $T u$ that is the image of another GP $u$ under a linear transformation $T$ acting on the sample paths of $u$ are well known, almost to the point of being folklore. However, these formulae are often used without rigorous attention to technical details, particularly when $T$ is an unbounded operator such as a differential operator, which is common in many modern applications. This note provides a self-contained proof of the claimed formulae for the case of a closed, densely-defined operator $T$ acting on the sample paths of a square-integrable (not necessarily Gaussian) stochastic process. Our proof technique relies upon Hille's theorem for the Bochner integral of a Banach-valued random variable.

Probability,Statistics Theory
Taylor expansions of solutions of stochastic partial differential equations

Arnulf Jentzen

DOI: https://doi.org/10.3934/dcdsb.2010.14.515

2009-04-15

Abstract:The solutions of parabolic and hyperbolic stochastic partial differential equations (SPDEs) driven by an infinite dimensional Brownian motion, which is a martingale, are in general not semi-martingales any more and therefore do not satisfy an Itô formula like the solutions of finite dimensional stochastic differential equations (SODEs). In particular, it is not possible to derive stochastic Taylor expansions as for the solutions of SODEs using an iterated application of the Itô formula. However, in this article we introduce Taylor expansions of solutions of SPDEs via an alternative approach, which avoids the need of an Itô formula. The main idea behind these Taylor expansions is to use first classical Taylor expansions for the nonlinear coefficients of the SPDE and then to insert recursively the mild presentation of the solution of the SPDE. The iteration of this idea allows us to derive stochastic Taylor expansions of arbitrarily high order. Combinatorial concepts of trees and woods provide a compact formulation of the Taylor expansions.

Numerical Analysis,Analysis of PDEs,Probability
Taylor Polynomials of Rational Functions

Aldo Conca,Simone Naldi,Giorgio Ottaviani,Bernd Sturmfels

2023-04-03

Abstract:A Taylor variety consists of all fixed order Taylor polynomials of rational functions, where the number of variables and degrees of numerators and denominators are fixed. In one variable, Taylor varieties are given by rank constraints on Hankel matrices. Inversion of the natural parametrization is known as Padé approximation. We study the dimension and defining ideals of Taylor varieties. Taylor hypersurfaces are interesting for projective geometry, since their Hessians tend to vanish. In three and more variables, there exist defective Taylor varieties whose dimension is smaller than the number of parameters. We explain this with Fröberg's Conjecture in commutative algebra.

Algebraic Geometry,Symbolic Computation,Commutative Algebra
Gaussian kernel expansion with basis functions uniformly bounded in $\mathcal{L}_{\infty}$

Mauro Bisiacco,Gianluigi Pillonetto

2024-10-02

Abstract:Kernel expansions are a topic of considerable interest in machine learning, also because of their relation to the so-called feature maps introduced in machine learning. Properties of the associated basis functions and weights (corresponding to eigenfunctions and eigenvalues in the Mercer setting) give insight into for example the structure of the associated reproducing kernel Hilbert space, the goodness of approximation schemes, the convergence rates and generalization properties of kernel machines. Recent work in the literature has derived some of these results by assuming uniformly bounded basis functions in $\mathcal{L}_\infty$. Motivated by this line of research, we investigate under this constraint all possible kernel expansions of the Gaussian kernel, one of the most widely used models in machine learning. Our main result is the construction on $\mathbb{R}^2$ of a Gaussian kernel expansion with weights in $\ell_p$ for any $p>1$. This result is optimal since we also prove that $p=1$ cannot be reached by the Gaussian kernel, nor by any of the other radial basis function kernels commonly used in the literature. A consequence for this kind of kernels is also the non-existence of Mercer expansions on $\mathbb{R}^2$, with respect to any finite measure, whose eigenfunctions all belong to a closed ball of $\mathcal{L}_\infty$.

Machine Learning
Generalized Taylor Series and Peano Kernel Theorem

Fatma Zürnacı‐Yetiş,Çetin Dişibüyük

DOI: https://doi.org/10.1002/mma.10616

2024-12-04

Mathematical Methods in the Applied Sciences

Abstract:As in the polynomial case, non‐polynomial divided differences can be viewed as a discrete analog of derivatives. This link between non‐polynomial divided differences and derivatives is defined by a generalization of the derivative operator. In this study, we obtain a generalization of Taylor series using the link between non‐polynomial divided differences and derivatives, and state generalized Taylor theorem. With the definition of a definite integral, the relation between the non‐polynomial divided difference and non‐polynomial B‐spline functions is given in terms of integration. Also, we derive a general form of the Peano kernel theorem based on a generalized Taylor expansion with the integral remainder. As in the polynomial case, it is shown that the non‐polynomial B‐splines are in fact the Peano kernels of non‐polynomial divided differences.MSC2020 Classification: 65D05, 65D07

mathematics, applied
Uniform approximation of common Gaussian process kernels using equispaced Fourier grids

Alex Barnett,Philip Greengard,Manas Rachh

DOI: https://doi.org/10.1016/j.acha.2024.101640

IF: 2.974

2024-02-01

Applied and Computational Harmonic Analysis

Abstract:The high efficiency of a recently proposed method for computing with Gaussian processes relies on expanding a (translationally invariant) covariance kernel into complex exponentials, with frequencies lying on a Cartesian equispaced grid. Here we provide rigorous error bounds for this approximation for two popular kernels—Matérn and squared exponential—in terms of the grid spacing and size. The kernel error bounds are uniform over a hypercube centered at the origin. Our tools include a split into aliasing and truncation errors, and bounds on sums of Gaussians or modified Bessel functions over various lattices. For the Matérn case, motivated by numerical study, we conjecture a stronger Frobenius-norm bound on the covariance matrix error for randomly-distributed data points. Lastly, we prove bounds on, and study numerically, the ill-conditioning of the linear systems arising in such regression problems.

mathematics, applied
Generalized Local Fractional Taylor's Formula with Local Fractional Derivative

Xiao-Jun Yang

DOI: https://doi.org/10.48550/arXiv.1106.2459

2012-06-29

Abstract:In the present paper, a generalized local Taylor formula with the local fractional derivatives (LFDs) is proposed based on the local fractional calculus (LFC). From the fractal geometry point of view, the theory of local fractional integrals and derivatives has been dealt with fractal and continuously non-differentiable functions, and has been successfully applied in engineering problems. It points out the proof of the generalized local fractional Taylor formula, and is devoted to the applications of the generalized local fractional Taylor formula to the generalized local fractional series and the approximation of functions. Finally, it is shown that local fractional Taylor series of the Mittag-Leffler type function is discussed.

Mathematical Physics
Taylor Series as Wide-sense Biorthogonal Wavelet Decomposition

H.M. de Oliveira,R.D. Lins

DOI: https://doi.org/10.48550/arXiv.1502.01570

2015-02-05

Abstract:Pointwise-supported generalized wavelets are introduced, based on Dirac, doublet and further derivatives of delta. A generalized biorthogonal analysis leads to standard Taylor series and new Dual-Taylor series that may be interpreted as Laurent Schwartz distributions. A Parseval-like identity is also derived for Taylor series, showing that Taylor series support an energy theorem. New representations for signals called derivagrams are introduced, which are similar to spectrograms. This approach corroborates the impact of wavelets in modern signal analysis.

Classical Analysis and ODEs,Information Theory,Mathematical Physics

The Taylor series of the Gaussian kernel

A probabilistic Taylor expansion with Gaussian processes

On the Gaussian Kinematic Formula of R. Adler and J. Taylor

Exponential Taylor Series

Taylor expansion for an operator function

On the Asymptotic Expansion of Bergman Kernel

Taylor's Law from Gaussian diffusions

Taylor series for the Askey-Wilson operator and classical summation formulas

Taylor's Formula for Generalized Weighted Fractional Derivatives with Nonsingular Kernels

Generalized Taylor's formula for power fractional derivatives

On Taylor series and Kapteyn series of the first and second type

On Taylor Series Expansion for Statistical Moments of Functions of Correlated Random Variables

Images of Gaussian and other stochastic processes under closed, densely-defined, unbounded linear operators

Taylor expansions of solutions of stochastic partial differential equations

Taylor Polynomials of Rational Functions

Gaussian kernel expansion with basis functions uniformly bounded in $\mathcal{L}_{\infty}$

Generalized Taylor Series and Peano Kernel Theorem

Uniform approximation of common Gaussian process kernels using equispaced Fourier grids

Generalized Local Fractional Taylor's Formula with Local Fractional Derivative

Taylor Series as Wide-sense Biorthogonal Wavelet Decomposition