Statistical network analysis

This framework will include results on both time-independent (static) and time-varying (dynamic) networks. You will use simple models as vehicles, aiming to quantify the uncertainties of statistical procedures’ outputs. The impact of this research will be beneficial to both the immediate academic community and practitioners in applied areas.

Networks are fundamental tools for representing relational data, which are ubiquitous across a wide range of application areas including public health, life science, social science, finance and cyber security, to name but a few.  

For instance in neuroscience, representing brain voxels and the functional connectivities in between as nodes and edges in networks facilitates the study of brain structures; in epidemiology, symbolising individuals and disease transmission incidences as nodes and edges enables decision makers to endorse effective policies. This abstraction calls for powerful mathematical tools, among which, statistics shines out when data sets are of massive scale.

In spite of the tremendous rate of growth of statistical network analysis, the theory of statistical testing procedures on networks is a largely uncharted territory and the status quo in studies of dependent networks relies heavily on complicated modelling. It is worth bearing in mind that

• testing is foundational for drawing meaningful conclusions from a statistical procedure
• dependence is the nature of networked data
• while complicated modelling methods are often preferred when the focus is prediction, they are limited when one is interested in understanding the data generating processes, and are notorious for their inflexibility and susceptibility to noise inflation.

Please contact Yi Yu for more details