Vol 3, No 2 (2019)

Approximation Theory & Meshless Methods

Authors: Umesh Patil, Depesh Deshpande, Ankur. Choudhury

Abstract: Approximation theory forms the mathematical backbone of many numerical and computational methods used to solve real-world problems where exact solutions are difficult or impossible to obtain. In recent decades, meshless (or meshfree) methods have emerged as a powerful alternative to classical mesh-based numerical techniques such as finite difference, finite element, and finite volume methods. These approaches rely heavily on approximation theory, particularly interpolation, regression, and functional approximation using scattered data. This paper presents a comprehensive review of approximation theory and its close connection with meshless methods. Fundamental concepts such as polynomial approximation, radial basis functions, moving least squares, and kernel-based approximations are discussed in detail. Various meshless methods, including the Element-Free Galerkin method, Smoothed Particle Hydrodynamics, and Local Petrov–Galerkin methods, are reviewed with emphasis on their theoretical foundations, advantages, and limitations. Applications in engineering, fluid dynamics, solid mechanics, and computational physics are also highlighted. The paper aims to provide a unified perspective that connects classical approximation theory with modern meshless numerical techniques, while also identifying current challenges and future research directions.

Keywords: Approximation theory, meshless methods, radial basis functions, moving least squares, numerical analysis, scattered data approximation

Full Issue

View or download the full issue PDF 104-114

Table of Contents