Detection of interacting variables for generalized linear models via neural networks
Open Access
- 1 November 2023
- journal article
- research article
- Published by Springer Nature in European Actuarial Journal
- Vol. 14 (2) , 551-580
Abstract
The quality of generalized linear models (GLMs), frequently used by insurance companies, depends on the choice of interacting variables. The search for interactions is time-consuming, especially for data sets with a large number of variables, depends much on expert judgement of actuaries, and often relies on visual performance indicators. Therefore, we present an approach to automating the process of finding interactions that should be added to GLMs to improve their predictive power. Our approach relies on neural networks and a model-specific interaction detection method, which is computationally faster than the traditionally used methods like Friedman’s H-Statistic or SHAP values. In numerical studies, we provide the results of our approach on artificially generated data as well as open-source data.Keywords
All Related Versions
Funding Information
- Technische Universität München
This publication has 18 references indexed in Scilit:
- Nesting Classical Actuarial Models into Neural NetworksSSRN Electronic Journal, 2019
- EDITORIAL: YES, WE CANN!ASTIN Bulletin, 2018
- Insights from Inside Neural NetworksSSRN Electronic Journal, 2018
- iml: An R package for Interpretable Machine LearningJournal of Open Source Software, 2018
- A data driven binning strategy for the construction of insurance tariff classesScandinavian Actuarial Journal, 2018
- Non-Life Insurance Pricing with Generalized Linear Models2010
- Predictive learning via rule ensemblesThe Annals of Applied Statistics, 2008
- Actuarial Modelling of Claim Counts2007
- Regression Shrinkage and Selection Via the LassoJournal of the Royal Statistical Society Series B: Statistical Methodology, 1996
- Generalized Linear ModelsJournal of the Royal Statistical Society. Series A (General), 1972