No Thumbnail Available
Details
- Supervisors
- Faculty
- Degree label
- Abstract
- The thesis is split into three main chapters. Chapter 1 Micro-modelling: In this chapter, we put ourselves into the shoes of one individual in the whole network. The goal is to check how his or her social contact behavior could influence on the infection of disease. For example, some friends may meet each other very regularly, say once a week while some others can be much various like sometimes several times a week and than loose contact of a couple of weeks. Even in the case where the average numbers of contact are the same for both types of contact, their probability of infect each other could still be quite different. We define a new type of stochastic dominance which is different from the classic ones. Chapter 2: Macro-modelling: In chapter 2, we use a more global vision for network diffusion. We are actually interested by the structure of social networks and its consequence on disease spreading. The forms of infection networks literally depend on the type of contacts that may contaminate the disease. For example the virus spread by sexual contact such as HIV will surely not share the same contact network with diseases that could spread through air such as influenza. The key quantity in this chapter is the reproductive ratio. We also define two different models (not sure if there exists exactly the same before) to simulate the spread of disease based on a given structure of contact matrix.Two real data set have also been used for the simulation. Chapter 3: Optimal vaccination strategy. Following the topic of chapter 2, we suppose the social contact network is perfectly known. We also make the assumption that we are only capable to vaccinate a part of the whole population. Therefore, we look for the most efficient measure to decide who should be vaccinated. The goal is to prevent the disease outbreak. We define a new quantity named as the transition capability measure based on the similarity of nodes in the network. We compare it with other measures such as degree and betweenness centrality on different contact networks.