Mercredi 8 mars 2017
Topological resonances on quantum graphs
Françoise Truc, (Université de Grenoble, Institut Fourier)
We consider metric graphs which consist of a finite graph with some
to some vertices. To this graph is associated a Laplacian using the
Kirchoff conditions. We describe some asymptotic properties of the resonances close to
the real axis. This is a joint work with Y. Colin de Verdière.
Classification of randomized reference models for temporal network data
Taro Takaguchi (National Institute of Information and Communications Technology, Tokyo)
When we have a network data, randomization techniques give us insights into its structural characteristics. The networks resampled after randomizations work as a baseline for a comparison with the empirical network (like a null model for statistical hypothesis testing). This randomization framework has successfully unveiled notable features of complex “static” networks. When the focal data is “temporal” networks, researchers have also proposed different randomization techniques. However, the addition of temporality makes the number of possible combinations of randomization procedures diverge, and thus choosing and designing suitable randomization techniques are nontrivial problems. As a first step towards developing a unified framework of randomization, we propose a taxonomy of existing randomization techniques, based on their methodological nature and their effects on network structure. We hope this work be a starting point for the development of the principled approach for the characterization of general time-varying network data based on randomizations.
This work is the result of a collaboration project with Laetitia Gauvin (ISI), Marton Karsai (ENS Lyon), Mikko Kivela (Aalto), Mathieu Genois (GESIS), Eugenio Valdano (Universitat Rovira i Virgili), and Christian Vestergaard (Institut Pasteur).