WAGNER WARPING OF GRAPH AND NETWORK REPRESENTATIONS
In the last decades, complex networks have emerged as a powerful tool for the modelling, analysis, and understanding of complex systems. Within them, functional networks have been at the foundations of the analysis of many real-world systems, the epitome possibly being the analysis of human brain dynamics. When a practitioner has to reconstruct a functional network representation, a fundamental question is what ought to be the links of such network, as these have to be inferred from observational data. Notably, much less attention has been devoted to a similar question involving nodes. In other words: could networks be created, in which nodes do not necessarily correspond to physical elements of the analysed system, but rather to combinations thereof? We here propose to explore this new form of network reconstruction, termed network warping; evaluate different computational strategies and objective metrics for its optimisation; and to assess its relevance in a set of biomedical and technological case studies.
Researchers
Massimiliano Zanin