On the Expected Discounted Penalty Function Using Physics-Informed Neural Network

We study the expected discounted penalty at ruin under a stochastic discount rate for the compound Poisson risk model with a threshold dividend strategy. The discount rate is modeled by a Poisson process and a standard Brownian motion. By applying the differentiation method and total expectation for...

Full description

Saved in:
Bibliographic Details
Published in:Journal of mathematics (Hidawi) 2023, Vol.2023, p.1-16
Main Authors: Wang, Jiayu, Wang, Houchun
Format: Article
Language:English
Subjects:
Actuarial science
Artificial intelligence
Composite materials
Discount rates
Discounts
Dividends
Formulas (mathematics)
Insurance companies
Laplace transforms
Neural networks
Partial differential equations
Penalty function
Physics
Poisson density functions
Random variables
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We study the expected discounted penalty at ruin under a stochastic discount rate for the compound Poisson risk model with a threshold dividend strategy. The discount rate is modeled by a Poisson process and a standard Brownian motion. By applying the differentiation method and total expectation formula, we obtain an integrodifferential equation for the expected discounted penalty function. From this integrodifferential equation, a renewal equation and an asymptotic formula satisfied by the expected discounted penalty function are derived. In order to solve the integrodifferential equation, we use a physics-informed neural network (PINN) for the first time in risk theory and obtain the numerical solutions of the expected discounted penalty function in some special cases of the penalty at ruin.
ISSN:2314-4629
2314-4785
DOI:10.1155/2023/9950023