Computational Mathematics and Computer Modeling with Applications (CMCMA)

Operational matrix formulation of Legendre LSSVR for Caputo-type integro-differential equations

https://doi.org/10.48308/CMCMA.4.2.1
Volume 4, Issue 2
September 2025
Pages 1-8
Computational Mathematics and Computer Modeling with Applications (CMCMA)

Document Type : Regular paper

Authors

Armin Ghasem Nejad 1
Alireza Afzal Aghaei 2
Sohrab Ali Yousefi 1
Kourosh Parand 3

1 Department of Applied Mathematics, Faculty of Mathematical Sciences, Shahid Beheshti University, Tehran, Iran.

2 Department of Computer and Data Sciences, Faculty of Mathematical Sciences, Shahid Beheshti University, Tehran, Tehran, Iran.

3 Department of Computer and Data Sciences, Shahid Beheshti University, Tehran, Iran

Abstract
This study develops a numerical framework for solving Caputo-type fractional integro-differential equations using a hybrid formulation that integrates Legendre polynomial operational matrices within the Least Squares Support Vector Regression (LSSVR) paradigm. The proposed method reformulates the governing equations by representing all differential, integral, and integral-kernel operators in matrix form, derived from the orthogonality and recurrence properties of Legendre bases. This leads to a fully vectorized constrained optimization problem, subsequently reduced to a symmetric linear system via Karush–Kuhn–Tucker conditions. The approach retains the spectral convergence characteristics of orthogonal polynomial methods while incorporating the regularization and flexibility of LSSVR. Numerical experiments on benchmark problems with known solutions demonstrate exponential accuracy and high numerical stability across varying polynomial degrees. This formulation accommodates both Fredholm and Volterra integral structures and directly incorporates Caputo initial conditions without reformulation. The method is compatible with higher-dimensional extensions and adaptable to various kernel approximations.

Keywords

Fractional Calculus
Operational Matrix
Legendre Polynomials
  • Receive Date 15 August 2025
  • Revise Date 22 August 2025
  • Accept Date 04 September 2025
Article View 11,412
PDF Download 11,535

Home

Editorial Board

Guide for Authors

Submit Manuscript

Contact Us

Generative AI policies

Copyright Policy

Licensing Policy

Plagiarism Policy

Self-archiving policies for authors

Open Access Statement