A Deep Neural Network Algorithm for Semilinear Elliptic PDEs with Applications in Insurance Mathematics

Research output: Contribution to journalArticleResearchpeer-review

Standard

A Deep Neural Network Algorithm for Semilinear Elliptic PDEs with Applications in Insurance Mathematics. / Kremsner, Stefan; Steinicke, Alexander; Szölgyenyi, Michaela.
In: Risks, Vol. 8, No. 4, 136, 09.12.2020, p. 1-18.

Research output: Contribution to journalArticleResearchpeer-review

Vancouver

Kremsner S, Steinicke A, Szölgyenyi M. A Deep Neural Network Algorithm for Semilinear Elliptic PDEs with Applications in Insurance Mathematics. Risks. 2020 Dec 9;8(4):1-18. 136. doi: 10.3390/risks8040136

Author

Bibtex - Download

@article{12f7ac296f7e41fa8a0f41f345003d9b,
title = "A Deep Neural Network Algorithm for Semilinear Elliptic PDEs with Applications in Insurance Mathematics",
abstract = "In insurance mathematics, optimal control problems over an infinite time horizon arise when computing risk measures. An example of such a risk measure is the expected discounted future dividend payments. In models which take multiple economic factors into account, this problem is high-dimensional. The solutions to such control problems correspond to solutions of deterministic semilinear (degenerate) elliptic partial differential equations. In the present paper we propose a novel deep neural network algorithm for solving such partial differential equations in high dimensions in order to be able to compute the proposed risk measure in a complex high-dimensional economic environment. The method is based on the correspondence of elliptic partial differential equations to backward stochastic differential equations with unbounded random terminal time. In particular, backward stochastic differential equations—which can be identified with solutions of elliptic partial differential equations—are approximated by means of deep neural networks.",
keywords = "backward stochastic differential equations, semilinear elliptic partial differential equations, stochastic optimal control, unbounded random terminal time, machine learning, deep neural networks",
author = "Stefan Kremsner and Alexander Steinicke and Michaela Sz{\"o}lgyenyi",
year = "2020",
month = dec,
day = "9",
doi = "10.3390/risks8040136",
language = "English",
volume = "8",
pages = "1--18",
journal = "Risks",
issn = "2227-9091",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "4",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - A Deep Neural Network Algorithm for Semilinear Elliptic PDEs with Applications in Insurance Mathematics

AU - Kremsner, Stefan

AU - Steinicke, Alexander

AU - Szölgyenyi, Michaela

PY - 2020/12/9

Y1 - 2020/12/9

N2 - In insurance mathematics, optimal control problems over an infinite time horizon arise when computing risk measures. An example of such a risk measure is the expected discounted future dividend payments. In models which take multiple economic factors into account, this problem is high-dimensional. The solutions to such control problems correspond to solutions of deterministic semilinear (degenerate) elliptic partial differential equations. In the present paper we propose a novel deep neural network algorithm for solving such partial differential equations in high dimensions in order to be able to compute the proposed risk measure in a complex high-dimensional economic environment. The method is based on the correspondence of elliptic partial differential equations to backward stochastic differential equations with unbounded random terminal time. In particular, backward stochastic differential equations—which can be identified with solutions of elliptic partial differential equations—are approximated by means of deep neural networks.

AB - In insurance mathematics, optimal control problems over an infinite time horizon arise when computing risk measures. An example of such a risk measure is the expected discounted future dividend payments. In models which take multiple economic factors into account, this problem is high-dimensional. The solutions to such control problems correspond to solutions of deterministic semilinear (degenerate) elliptic partial differential equations. In the present paper we propose a novel deep neural network algorithm for solving such partial differential equations in high dimensions in order to be able to compute the proposed risk measure in a complex high-dimensional economic environment. The method is based on the correspondence of elliptic partial differential equations to backward stochastic differential equations with unbounded random terminal time. In particular, backward stochastic differential equations—which can be identified with solutions of elliptic partial differential equations—are approximated by means of deep neural networks.

KW - backward stochastic differential equations

KW - semilinear elliptic partial differential equations

KW - stochastic optimal control

KW - unbounded random terminal time

KW - machine learning

KW - deep neural networks

UR - http://www.scopus.com/inward/record.url?scp=85097538081&partnerID=8YFLogxK

U2 - 10.3390/risks8040136

DO - 10.3390/risks8040136

M3 - Article

VL - 8

SP - 1

EP - 18

JO - Risks

JF - Risks

SN - 2227-9091

IS - 4

M1 - 136

ER -