| KEYWORD |
Network Science, Modeling of Dynamical Systems
Thesis abroad
keywords COMPLEX NETWORKS, EPIDEMIC MODELING, IDENTIFICATION, NETWORK SCIENCE, TIME-VARYING NETWORKS
Reference persons GIUSEPPE CARLO CALAFIORE, ALESSANDRO RIZZO
Research Groups SYSTEMS AND DATA SCIENCE - SDS
Thesis type SIMULATIVA
Description Mathematical models of infectious disease spread are potent tools for the management of dangerous outbreaks. These models can form a basis for planning and implementing vaccination strategies, evaluating the risks and benefits of travel bans, and improving the effectiveness of prophylaxis campaigns. However traditional modeling approaches do not fully capture the national and international mobility characteristic of modern society, where contacts do not remain geographically confined to the area of the initial outbreak, and an infection may jump thousands of miles in a single day. This project will advance fundamental understanding of dynamical systems evolving on reconfigurable networks, in which the subsystems and the network connections change on comparable time-scales. The resulting mathematical framework will enable a new class of predictive models of infectious disease spread. These models will aid in safeguarding uninfected populations and in mitigating impact on afflicted nations, even when, as in the case of Ebola Virus Disease, no therapeutic protocol is available. More broadly, the underlying theoretical advances are expected to transform the analysis, design, and control of dynamical systems on rapidly reconfiguring networks.
This thesis seeks to advance the field of dynamical systems and complex networks toward tractable mathematical models of infectious disease epidemics. Specifically, this thesis will establish a theoretical framework for the study of the concurrent evolution of the dynamics of infectious diseases and the formation of the network of contacts through which they spread. The framework will be based on the notion of activity-driven networks, which can be effectively utilized to model contact processes that evolve over time-varying networks across a range of time-scales. This modeling paradigm contrasts that of traditional connectivity-driven networks, where links between nodes have a long life span, resulting in the separation between the time-scales of the dynamics of the network connections and the process evolution. In the thesis activity we will seek to understand the effect of non-ideal containment procedures on the spread of infectious disease through the systematic analysis of global and local network features; devise strategies for community detection in time-varying networks, toward identifying untraced contacts that are critical for disease spreading and of great public concern; and establish model-based optimization strategies to prioritize contact tracing procedures toward improving the effectiveness and outcomes of control interventions. A reference paper is attached.
See also rp_jtb_2016.pdf
Required skills Dynamical Systems Theory, Foundations of graph theory will be a plus. Ability to code discrete-time systems.
Notes There is the opportunity to carry out part of the thesis at the New York University Tandon School of Engineering, under the co-mentoring of Prof. Maurizio Porfiri.
Deadline 10/02/2017
PROPONI LA TUA CANDIDATURA