Adaptive approaches for predictive modelling of Complex Multilayered Networks
Nowadays, we have the possibility to process ‘big data’ (gathered in computer systems) about the interactions and activities of millions of individuals. It represents an increasingly important yet underutilised resource because due to the scale, complexity and dynamics, Complex Multilayer Networks (CSN) extracted from this data are extremely difficult to analyse. So far, there is no coherent and comprehensive approach to analyse multilayer networks which represent multiple types of relationships that can exist between nodes. To overcome the limitations of existing analytical techniques, the main goal of this project is to develop and evaluate an adaptive framework for predictive analysis of complex multilayer networks.
Most of the research within real-world complex networks prediction is limited to the problem of link prediction. Out of those, majority address only the cases where links are added to the network. Some models, such as preferential attachment, take into account adding new nodes but none of them cope with the combined problem of predicting (a) the probability of forming/fading of a connection, (b) changing number of nodes within the network, and (c) changing features of both nodes and relations. This is a big challenge that this project will address by providing a predictive platform that copes with all types of predictions within CSN.
Another very challenging and also not yet addressed problem is “dynamics of dynamics” i.e. how the dynamics evolves over time depending on e.g. in what state (growth, maturity, death) the network is. This means that the predictive models developed within this project will need to adapt to the changing nature of input data.
This project will result in foundational contributions to modelling of complex multilayer networks which are currently missing. There is a need for reshaping the landscape of predictive analytics in complex networks and if you think that this is something you want to do then please apply for this PhD scholarship.
Supervisors: Dr Katarzyna Musial-Gabrys (University of Technology Sydney)
Deadline for Applications: Available now until filled
Before you apply and for more info, please contact Katarzyna Musial-Gabrys at katarzyna.musial- [Email Address Removed]
Stipend value $27,082 pa (tax exempt); only full-time, 3 years.