A1434
Title: Model selection for extremal dependence structures using deep learning: Application to environmental data
Authors: Veronique Maume-Deschamps - University Lyon 1 (France)
Pierre Ribereau - Université Lyon 1 (France) [presenting]
Manaf Ahmed - University of Mosul (Iraq)
Abstract: The purpose is to introduce a new methodology for extreme spatial dependence structure selection. It is based on deep learning techniques, specifically convolutional neural networks (CNNs). Two schemes are considered: in the first scheme, the matching probability is evaluated through a single CNN, while in the second scheme, a hierarchical procedure is proposed: a first CNN is used to select a max-stable model, then another network allows to select the most adapted covariance function, according to the selected max-stable model. This model selection approach demonstrates good performance on simulations. On the contrary, the composite likelihood information criterion (CLIC) faces issues in selecting the correct model. Both schemes are applied to a dataset of 2m air temperature over Iraq land, CNNs are trained on dependence structures summarized by the concurrence probability.
