Leiden University Student Repository

Claus, Benjamin (2024)

Reciprocation in scale-invariant network models

Master thesis | Physics (MSc)

In this research, a recently proposed renormalization group approach for networks to the case of random directed graphs is being generalized: we present a scale-invariant description of directed networks containing reciprocated edges. This allows us to neglect several strong assumptions that are currently necessary to renormalize directed networks such as financial transaction networks. As an application, a model of ING’s transaction data has been derived across multiple coarse-grained partitions. In this article we provide detailed information on how this particular model has been structured and how its parameters are obtained. We show how we can use this model to determine the expected cumulative degree and weight distributions of ING’s transaction network across multiple coarse-grained partitions of the network which we will compare to the empirical degree and weight distributions, respectively.

Author

Claus, Benjamin

Faculty

Faculty of Science

Specialisation

Theoretical physics

Supervisors

D. Garlaschelli

ECTS Credits

48

Language

en