Εντοπίστηκε ένα σφάλμα στη λειτουργία της ΠΥΞΙΔΑΣ όταν χρησιμοποιείται μέσω του προγράμματος περιήγησης Safari. Μέχρι να αποκατασταθεί το πρόβλημα, προτείνουμε τη χρήση εναλλακτικού browser όπως ο Chrome ή ο Firefox. A bug has been identified in the operation of the PYXIDA platform when accessed via the Safari browser. Until the problem is resolved, we recommend using an alternative browser such as Chrome or Firefox.
Λογότυπο αποθετηρίου

Διδακτορικές διατριβές

Μόνιμο URI για αυτήν τη συλλογήhttps://pyxida.aueb.gr/handle/123456789/63

Περιήγηση

Πλοήγηση Διδακτορικές διατριβές ανά Θέμα "Aξία της πληροφορίας"

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
    Essays on epidemic models and their statistical analysis
    (09/22/2021) Chatzilena, Anastasia; Χατζηλένα, Αναστασία; Athens University of Economics and Business, Department of Economics; Demiris, Nikolaos; Genakos, Christos; Sypsa, Vana; Baguelin, Marc; Kalogeropoulos, Kostas; Kypraios, Theodore; Arvanitis, Stylianos
    Public health related decisions concerning infectious diseases are characterized by the use of complex mathematical models in order to understand the dynamics of infectious diseases and design intervention strategies. Efficient modelling and inference procedures for learning the model parameters from data are of central interest. In this thesis, a comprehensive review of two new and efficient statistical machine learning methods, namely Hamiltonian Monte Carlo and Variational Inference, as implemented in the freely available Stan software, was carried out. We explored how Stan could be used to fit a class of epidemic models based upon systems of ordinary differential equations and demonstrated its potential in an application to real data. In the light of the COVID-19 pandemic, this thesis revolved around model-based approaches to estimate the transmissibility of SARS-CoV2, focusing on two different classes of epidemic models. Shortcomings in global epidemiological surveillance, led to the use of indirect estimations of infections, through deaths. This approach was adopted to fit an extension of the deterministic SEIR (susceptible-exposed-infected-recovered) compartmental model, where the transmission rate is a diffusion process, allowing to reveal both the effect of control strategies and the changes in individuals behaviour. We proceeded with a suitably tailored chain-binomial epidemic model which was later extended to include population heterogeneity, introducing contact uncertainty into the inference structure in a highly hierarchical setting, trying to reveal the age distribution of infections through aggregate deaths. In the main, careful consideration of data combined with the use of contemporary developments in statistics, can be an essential tool for advanced analysis based on realistically complex models.