Full text
| |
| Author(s): |
Paulo Cesar Ventura da Silva
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Carlos. |
| Institution: | Universidade de São Paulo (USP). Instituto de Física de São Carlos (IFSC/BT) |
| Defense date: | 2022-01-20 |
| Examining board members: |
Francisco Aparecido Rodrigues;
Marcus Aloizio Martinez de Aguiar;
Henrique Ferraz de Arruda;
Silvio da Costa Ferreira Junior;
Angélica Sousa da Mata
|
| Advisor: | Francisco Aparecido Rodrigues |
| Abstract | |
Epidemics have always been the cause of significant economic and life losses. Among the multiple developments of science against infectious diseases, mathematical and computational models are increasingly important, especially after the development of complexity and network science. The study and forecasting of epidemics are highly influenced by behaviors of host populations, which are unpredictable and difficult to incorporate. In this thesis, we contribute to the substantial volume of works dedicated to solving this problem. First, we approach the coupling between disease and information spreading on multiplex networks. We propose a rumor-like model for the information and a flexible ratio of time scales between disease and information. We show that increasing the time scale of the information reduces the disease prevalence. We also show that stiflers may cause an increase in both infection and information levels. This problem is then further studied with more general models of asymmetrical interaction between contagion phenomena. We show that our previous results on the time scale depend on the form of interaction between the spreading processes. We also study in more depth these models, numerically describing their transient oscillations and deriving analytical expressions for their steady states and phase transitions. We then switch to systems with host mobility, which is an important ingredient of epidemic spreading. We first develop an individual-based mobility and epidemic model, into which behavioral responses are incorporated as an avoidance of infectious hosts. We show how this reduces the disease spreading in different regimes of our model. In particular, for when mobility evolves much faster than epidemics, we derive a semi-analytical approach to describe the model´s bifurcation diagrams, verifying the existence of a bistable region and relating the dynamics to some metrics of the underlying networks. Finally, we move to larger scales and describe mobility as net flows between homogeneous populations. In this metapopulation scheme, we propose a model for behavioral responses that directly reduce the disease reproduction number. We show that our model can generate different outbreak sizes between subpopulations. Then we use it to compare strategies in which each subpopulation responds independently (locally), or the whole population follows the same response curve (globally). We show which strategy is more effective in different scenarios, for both a random geometric graph and two metapopulations constructed from real data. With the variety of topics that we approached, we hope to contribute significantly to the problem of disease-behavior coupling. (AU) | |
| FAPESP's process: | 16/24555-0 - Propagation of information in multilayer networks |
| Grantee: | Paulo Cesar Ventura da Silva |
| Support Opportunities: | Scholarships in Brazil - Doctorate |