The spread of epidemics, especially when coupled with human behavior dynamics, presents a complex challenge for accurate modeling and prediction. Traditionally, epidemic models rely on mobility matrices to represent the movement of individuals between locations, mimicking the spread of particles through a network. The disadvantage of this approach is that the population within each spatial patch changes over time. In this thesis, we explore an alternative approach to improve epidemic models by: (i) incorporating multiple human behavioral factors happening simultaneously, at different scales; (ii) integrating the concept of force of infection, which accounts for both spatial proximity and infection risk, including the effects of human mobility dynamics without changing the patch population. To evaluate the efficacy of this approach, we perform a comparative analysis between traditional epidemic models based on particle diffusion and our proposed model that incorporates force of infection. By examining critical behaviors, from the study of the critical point that separates absorbing and active regimes, under various scenarios, our goal is to elucidate the advantages and limitations of each approach. Specifically, in addition to the classical epidemic threshold posed by the condition of the reproduction number $\mathcal{R}_0$ equal to $1$, in metapopulation networks a new global invasion threshold $\mathcal{R}_\ast$ is found and its dependence on the interaction between patches is assessed. In parallel to the classical reaction-diffusion approach, the force-of-infection paradigm is introduced and used to check the consistency of the previous results. The main advantage of this new approach is the effective mapping of the flows of individuals moving between subpopulations, avoiding the need to assume a Markovian mobility. The knowledge acquired is then used to modify and improve a pre-existing model for seasonal influenza in Italy. For practical purposes, we will use Particle Swarm Optimization (PSO) techniques for fine-tuning model parameters to better capture the intricate interplay between disease dynamics and human behavior patterns and allow for analysis on their relevance to the predictions of the model. The final purpose of this model is to provide a new independent estimate for the national seasonal influenza forecasting service, Influcast, to improve the reliability of the estimates already available. In general, this research contributes to the advancement of epidemic modeling by providing insight into the complex interdependencies between disease dynamics and human behavior in terms of mobility and group contact dynamics, thereby enhancing our ability to develop more accurate and robust strategies for epidemic control and mitigation.
The spread of epidemics, especially when coupled with human behavior dynamics, presents a complex challenge for accurate modeling and prediction. Traditionally, epidemic models rely on mobility matrices to represent the movement of individuals between locations, mimicking the spread of particles through a network. The disadvantage of this approach is that the population within each spatial patch changes over time. In this thesis, we explore an alternative approach to improve epidemic models by: (i) incorporating multiple human behavioral factors happening simultaneously, at different scales; (ii) integrating the concept of force of infection, which accounts for both spatial proximity and infection risk, including the effects of human mobility dynamics without changing the patch population. To evaluate the efficacy of this approach, we perform a comparative analysis between traditional epidemic models based on particle diffusion and our proposed model that incorporates force of infection. By examining critical behaviors, from the study of the critical point that separates absorbing and active regimes, under various scenarios, our goal is to elucidate the advantages and limitations of each approach. Specifically, in addition to the classical epidemic threshold posed by the condition of the reproduction number $\mathcal{R}_0$ equal to $1$, in metapopulation networks a new global invasion threshold $\mathcal{R}_\ast$ is found and its dependence on the interaction between patches is assessed. In parallel to the classical reaction-diffusion approach, the force-of-infection paradigm is introduced and used to check the consistency of the previous results. The main advantage of this new approach is the effective mapping of the flows of individuals moving between subpopulations, avoiding the need to assume a Markovian mobility. The knowledge acquired is then used to modify and improve a pre-existing model for seasonal influenza in Italy. For practical purposes, we will use Particle Swarm Optimization (PSO) techniques for fine-tuning model parameters to better capture the intricate interplay between disease dynamics and human behavior patterns and allow for analysis on their relevance to the predictions of the model. The final purpose of this model is to provide a new independent estimate for the national seasonal influenza forecasting service, Influcast, to improve the reliability of the estimates already available. In general, this research contributes to the advancement of epidemic modeling by providing insight into the complex interdependencies between disease dynamics and human behavior in terms of mobility and group contact dynamics, thereby enhancing our ability to develop more accurate and robust strategies for epidemic control and mitigation.
Critical behaviour in multiscale epidemic models based on force of infection
BERTOLA, TOMMASO
2023/2024
Abstract
The spread of epidemics, especially when coupled with human behavior dynamics, presents a complex challenge for accurate modeling and prediction. Traditionally, epidemic models rely on mobility matrices to represent the movement of individuals between locations, mimicking the spread of particles through a network. The disadvantage of this approach is that the population within each spatial patch changes over time. In this thesis, we explore an alternative approach to improve epidemic models by: (i) incorporating multiple human behavioral factors happening simultaneously, at different scales; (ii) integrating the concept of force of infection, which accounts for both spatial proximity and infection risk, including the effects of human mobility dynamics without changing the patch population. To evaluate the efficacy of this approach, we perform a comparative analysis between traditional epidemic models based on particle diffusion and our proposed model that incorporates force of infection. By examining critical behaviors, from the study of the critical point that separates absorbing and active regimes, under various scenarios, our goal is to elucidate the advantages and limitations of each approach. Specifically, in addition to the classical epidemic threshold posed by the condition of the reproduction number $\mathcal{R}_0$ equal to $1$, in metapopulation networks a new global invasion threshold $\mathcal{R}_\ast$ is found and its dependence on the interaction between patches is assessed. In parallel to the classical reaction-diffusion approach, the force-of-infection paradigm is introduced and used to check the consistency of the previous results. The main advantage of this new approach is the effective mapping of the flows of individuals moving between subpopulations, avoiding the need to assume a Markovian mobility. The knowledge acquired is then used to modify and improve a pre-existing model for seasonal influenza in Italy. For practical purposes, we will use Particle Swarm Optimization (PSO) techniques for fine-tuning model parameters to better capture the intricate interplay between disease dynamics and human behavior patterns and allow for analysis on their relevance to the predictions of the model. The final purpose of this model is to provide a new independent estimate for the national seasonal influenza forecasting service, Influcast, to improve the reliability of the estimates already available. In general, this research contributes to the advancement of epidemic modeling by providing insight into the complex interdependencies between disease dynamics and human behavior in terms of mobility and group contact dynamics, thereby enhancing our ability to develop more accurate and robust strategies for epidemic control and mitigation.File | Dimensione | Formato | |
---|---|---|---|
Bertola_Tommaso.pdf
accesso aperto
Dimensione
3.69 MB
Formato
Adobe PDF
|
3.69 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/70806