Mean-square convergence of numerical methods for random ordinary differential equations (Q2021772)

This paper analyses the mean square convergence of numerical methods for random ordinary differential equations. A relation between global and local order of one step methods is proved. Global mean square convergence rates are investigated on a variety of problems (including those based on fractional Brownian motion) for Taylor methods, Affine-Taylor methods and Ito-Taylor methods, which are confirmed through numerical experiments.