Showing 1–1 of 1 results for author: Sawai, T

An efficient weak Euler-Maruyama type approximation scheme of very high dimensional SDEs by orthogonal random variables

Authors: Jirô Akahori, Masahiro Kinuya, Takashi Sawai, Tomooki Yuasa

Abstract: We will introduce Euler-Maruyama approximations given by an orthogonal system in $L^{2}[0,1]$ for high dimensional SDEs, which could be finite dimensional approximations of SPDEs. In general, the higher the dimension is, the more one needs to generate uniform random numbers at every time step in numerical simulation. The schemes proposed in this paper, in contrast, can deal with this problem by ge… ▽ More We will introduce Euler-Maruyama approximations given by an orthogonal system in $L^{2}[0,1]$ for high dimensional SDEs, which could be finite dimensional approximations of SPDEs. In general, the higher the dimension is, the more one needs to generate uniform random numbers at every time step in numerical simulation. The schemes proposed in this paper, in contrast, can deal with this problem by generating very few uniform random numbers at every time step. The schemes save time in the simulation of very high dimensional SDEs. In particular, we conclude that an Euler-Maruyama approximation based on the Walsh system is efficient in high dimensions. △ Less

Submitted 5 April, 2021; v1 submitted 15 January, 2020; originally announced January 2020.

Comments: 25 pages, 16 figures

MSC Class: 60H35; 65C05; 60H10