Code supporting "Hybrid Extreme Value and Deep Learning Models for Tail-Risk Analysis of COVID-19 Dynamics Across Nine Island Regions" in Scientific Reports

Résumé fourni par la source

This repository contains the computational code supporting the manuscript “Hybrid Extreme Value and Deep Learning Models for Tail-Risk Analysis of COVID-19 Dynamics Across Nine Island Regions”. The code implements two complementary frameworks for modelling extreme COVID-19 dynamics across nine island regions: (i) a deep-learning–Extreme Value Theory framework incorporating Generalized Pareto Distribution modelling of extreme residuals and return-level estimation, and (ii) an ARMA-(E)GARCH–Extreme Value Mixture Model framework incorporating conditional volatility, GPD tail modelling and case-scale return-level estimation. The archive includes Python and R implementations required to reproduce the principal modelling, diagnostic and return-level analyses reported in the study.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.