Neural Network Embedding of the Over-Dispersed Poisson Reserving Model

30 Pages Posted: 14 Dec 2018

See all articles by Andrea Gabrielli

Andrea Gabrielli

RiskLab, ETH Zurich

Ronald Richman

AIG South Africa

Mario V. Wuthrich

RiskLab, ETH Zurich

Date Written: November 21, 2018

Abstract

The main idea of this paper is to embed a classical actuarial regression model into a neural network architecture. This nesting allows us to learn model structure beyond the classical actuarial regression model if we use as starting point of the neural network calibration exactly the classical actuarial model. Such models can be fitted efficiently which allows us to explore bootstrap methods for prediction uncertainty. As an explicit example we consider the cross-classified over-dispersed Poisson model for general insurance claims reserving. We demonstrate how this model can be improved by neural network features.

Keywords: cross-classified over-dispersed Poisson model, neural network, model blending, nested models, learning across portfolios, claims reserving in insurance, chain-ladder reserves, mean square error of prediction

JEL Classification: G22, C02, C13, C14, C15, C45, C50

Suggested Citation

Gabrielli, Andrea and Richman, Ronald and Wuthrich, Mario V., Neural Network Embedding of the Over-Dispersed Poisson Reserving Model (November 21, 2018). Available at SSRN: https://ssrn.com/abstract=3288454 or http://dx.doi.org/10.2139/ssrn.3288454

Andrea Gabrielli

RiskLab, ETH Zurich ( email )

Department of Mathematics
Ramistrasse 101
Zurich, 8092
Switzerland

Ronald Richman

AIG South Africa ( email )

No Address Available

Mario V. Wuthrich (Contact Author)

RiskLab, ETH Zurich ( email )

Department of Mathematics
Ramistrasse 101
Zurich, 8092
Switzerland

Register to save articles to
your library

Register

Paper statistics

Downloads
308
Abstract Views
920
rank
97,377
PlumX Metrics