CFE-CMStatistics 2025: Start Registration
View Submission - CFE-CMStatistics 2025
A0761
Title: Nested Archimedean copula spatiotemporal approach for hail claim hazard Authors:  Melina Mailhot - Concordia (Canada) [presenting]
Nahid Sadr - Université de Sherbrooke (Canada)
Klaus Herrmann - Sherbrooke University (Canada)
Abstract: The objective of this project is to model hail claim frequencies for a specific province in Canada. A nested Archimedean copula model is used, with zero-inflated negative binomial GLM margins. Based on theoretical results related to the distortion of copulas, we are able to adapt the spatiotemporal model, which is compared to well-known models, such as INLA and models using spatiotemporal stochastic processes. It turns out the methodology, using distorted copulas, outperforms existing techniques.