Department of Econometrics & OR
VU University Amsterdam
My name is Bernd Heidergott. My background is Operations Research.
Since several years I am working on social networks where I am particularly interested in microscopic network models, i.e., models were the individual agents and their behavior are described by a fully characterized dynamic.
I am fascinated by the centrifugal forces captured in social networks. These forces seem to anticipate the near future of the unregulated network. Cluster identification, cluster categorization and (optimal) interventions in cluster-structures have my interest.
Social networks offer a variety of problems where the explicatory power of distributional limits is limited. Understanding the relations between the network behavior on the distributional level and the path-wise network behavior is a mathematically challenging topic (limits may occur backwards in time, the network may be stationary but non-ergodic, network exhibits no drifts etc.) Developing a multi-agent based Markov chain model to investigate these phenomena and analyze their relationship has my interest.
Data collection for analysis of social interactions is confronted with restrictive privacy regulations. I am interested in developing mathematical-based resampling methods that allow to use privacy-sensitive data so that the narrative of the data is captured but no individual interaction activities can be traced back.