Approximate Bayesian Computation and compartmental metapopulation model to describe COVID-19 in Finland
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Perustieteiden korkeakoulu |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2021-12-13
Department
Major/Subject
Machine Learning, Data Science and Artificial Intelligence
Mcode
SCI3044
Degree programme
Master’s Programme in Computer, Communication and Information Sciences
Language
en
Pages
72
Series
Abstract
At the beginning of the Coronavirus Disease 2019 (COVID-19) pandemic, when there were no vaccines available, a considerable number of governments around the world imposed mobility restrictions to control the spread of the disease. Mobility restrictions are part of the so-called Non-Pharmaceutical Interventions (NPIs). NPIs are measures taken by individuals or mandated by governments to prevent the spread of the disease. Some of the most notable NPIs include social distancing, use of face masks, remote working, closure of non-essential businesses, and travel restrictions. In Finland, these and other NPIs were essential to control the disease during the first wave of the epidemic, from March 2020 to May 2020. The aim of this thesis is to use a compartmental metapopulation model to study the COVID-19 epidemic and the role of NPIs in Finland. The model is informed with mobility data from a telecommunications operator and calibrated to describe the epidemic during the first wave. Furthermore, the model is calibrated using Approximate Bayesian Computing (ABC), a set of algorithms that make it possible to infer model parameters without an explicit function for the likelihood of data being produced by the model. The purpose of the model is to analyze the effect of NPIs in Finland through counterfactual scenarios. Counterfactual scenarios affect aspects of the calibrated model to present alternate scenarios of the original situation. This thesis uses two classes of counterfactual scenarios based on mobility data. The first class reduces or increases the number of contacts made inside the region. Similarly, the second class increases or reduces the cross-border traffic between the regions in Finland. The analysis performed in this thesis confirms that NPIs applied in Finland were crucial to reducing the incidence in Finland during the first wave. Moreover, NPIs which prevent contacts inside regions, such as social distancing, have a more significant impact on reducing the incidence than those that limit traffic between the regions. Finally, the results of this analysis, i.e., the observed effect of NPIs and the software package developed for calibrating the model through ABC, could further analyze the situation of COVID-19 or future epidemics.Description
Supervisor
Kivelä, MikkoThesis advisor
Leskelä, LasseAla-Nissilä, Tapio
Keywords
approxmate Bayesian Computation, epidemic modeling, coronavirus, non-pharmaceutical interventions, COVID-19, counterfactual scenarios