Strong convergence of an explicit numerical approximation for n-dimensional superlinear SDEs with positive solutions

Cai, Yongmei and Guo, Qian and Mao, Xuerong (2024) Strong convergence of an explicit numerical approximation for n-dimensional superlinear SDEs with positive solutions. Mathematics and Computers in Simulation, 216. pp. 198-212. ISSN 0378-4754 (https://doi.org/10.1016/j.matcom.2023.09.011)

[thumbnail of Cai_etal_MCS_2024_Strong_convergence_of_an_explicit_numerical_approximation_for_n_dimensional_superlinear_SDEs_with_positive_solutions]
Preview
Text. Filename: Cai_etal_MCS_2024_Strong_convergence_of_an_explicit_numerical_approximation_for_n_dimensional_superlinear_SDEs_with_positive_solutions.pdf
Accepted Author Manuscript
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (1MB)| Preview

Abstract

For a stochastic differential equation (SDE) with a unique positive solution, a rational numerical method is expected to be structure preserving. However, most existing methods are not, as far as we know. Some characteristics of the SDE models including the multi-dimension and super-linearity make it even more challenging. In this work, we fill the gap by proposing an explicit numerical method which is not only structure preserving but also cost effective. The strong convergence framework is set up by moment convergence analysis. We use the Lotka–Volterra system to elaborate our theory, nevertheless, the method works for a wide range of multi-dimensional superlinear SDE models.

ORCID iDs

Cai, Yongmei, Guo, Qian and Mao, Xuerong ORCID logoORCID: https://orcid.org/0000-0002-6768-9864;