Λογότυπο αποθετηρίου

Πλοήγηση ανά Συγγραφέα "Constantinou, Vasilia"

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω
Φιλτράρετε τα αποτελέσματα πληκτρολογώντας τα πρώτα γράμματα
Τώρα δείχνει 1 - 1 από 1
  • Μικρογραφία εικόνας
    Τεκμήριο
    Actuarial risk assessment: modeling COVID-19 mortality in Europe
    (2025-10-24) Constantinou, Vasilia; Κωνσταντίνου, Βασιλεία; Pedeli, Xanthi; Pateras, Konstantinos; Besbeas, Panagiotis
    This thesis models weekly COVID-19 mortality across 20 European countries using a spline-based Negative Binomial Generalized Linear Model (NB-GLM). The model incorporates B-splines for time, country fixed effects and key epidemiological, demographic and healthcare predictors, explaining 78% of the variation in mortality. New cases, ICU occupancy and lagged deaths emerged as strong predictors, reflecting both infection trends and healthcare strain. Demographic factors, particularly the proportion of elderly population and diabetes prevalence, were linked to higher baseline mortality. Vaccination effects were wave-dependent, with a protective impact in wave 2 but more complex dynamics in wave 3 due to timing and interactions with other factors. Country-specific effects showed substantial geographic disparities, with Bulgaria and Romania exhibiting higher mortality than countries like Italy and Portugal. Model diagnostics revealed mild residual autocorrelation and slight underestimation in extreme mortality weeks. For actuarial applications, the results support dynamic mortality loadings, wave and country-specific adjustments and the use of epidemiological indicators as early-warning signals for pricing and reserving. The framework provides a flexible, interpretable approach to pandemic mortality modeling, offering insights for both COVID-19 risk management and broader actuarial applications to public health crises.