Core Concepts
This paper introduces a theoretical framework for investigating analytic maps from finite discrete data, elucidating the mathematical machinery underlying polynomial approximation with least-squares in multivariate situations.
Abstract
The key insights and highlights of the content are:
The paper proposes an approach that does not directly approximate the analytic map itself, but instead approximates the push-forward on the space of locally analytic functionals induced by the analytic map. This enables a theoretical handling of approximation with least-squares using multivariate polynomials in more general situations.
The paper establishes a methodology for appropriate finite-dimensional approximation of the push-forward from finite discrete data, through the theory of the Fourier–Borel transform and the Fock space. It proves a rigorous convergence result with a convergence rate.
As an application, the paper proves that it is not the least-squares polynomial, but the polynomial obtained by truncating its higher-degree terms, that approximates analytic functions and allows for approximation beyond the support of the data distribution.
The paper shows that the finite-dimensional approximation of the push-forward enables the application of linear algebraic operations. Utilizing this, it proves the convergence of a method for approximating an analytic vector field from finite data of the flow map of an ordinary differential equation.
The paper provides a general theoretical framework for estimating analytic maps from finite discrete data, shedding light on the underlying machinery of polynomial approximation with least-squares in multivariate situations.
Quotes
"Our approach is to consider the push-forward on the space of locally analytic functionals, instead of directly handling the analytic map itself."
"One advantage of our theory is that it enables us to apply linear algebraic operations to the finite-dimensional approximation of the push-forward."