Control Variate Method for Deep BSDE Solver Using Weak Approximation
Yoshifumi Tsuchida ()
Additional contact information
Yoshifumi Tsuchida: Asset Management One Co., Ltd.
Asia-Pacific Financial Markets, 2023, vol. 30, issue 2, No 1, 273-296
Abstract:
Abstract The paper develops a new deep learning based scheme for solving high-dimensional nonlinear forward-backward stochastic differential equations (FBSDE) and associated partial differential equations. Firstly, the original BSDE is split into the linear dominant BSDE part and the nonlinear residual BSDE part. Then the linear BSDE part is approximated with high accuracy using a weak approximation technique. To approximate the nonlinear BSDE part, Deep BSDE solver is applied with asymptotic expansions which work as control variates. A sharp error estimate provides how the new scheme improves the original Deep BSDE method. Numerical experiments for high-dimensional nonlinear models show the validity and the effectiveness of the new scheme in financial application.
Keywords: Asymptotic expansion; Backward stochastic differential equation; Control variate method; Deep BSDE solver; Deep learning; Weak approximation (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations:
Downloads: (external link)
http://link.springer.com/10.1007/s10690-022-09374-8 Abstract (text/html)
Access to the full text of the articles in this series is restricted.
Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
HTML/Text
Persistent link: https://EconPapers.repec.org/RePEc:kap:apfinm:v:30:y:2023:i:2:d:10.1007_s10690-022-09374-8
Ordering information: This journal article can be ordered from
http://www.springer.com/finance/journal/10690/PS2
DOI: 10.1007/s10690-022-09374-8
Access Statistics for this article
Asia-Pacific Financial Markets is currently edited by Jiro Akahori
More articles in Asia-Pacific Financial Markets from Springer, Japanese Association of Financial Economics and Engineering
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().