Econometric Analysis of Network Data

Accounting for the relational nature of many social and economic activities, there has been an immense increase of interest in network and graph-based data. In particular, more and more datasets with natural graph theoretic (i.e., network) structure are increasingly available to researchers. In turn, a strong increase of system-level perspectives in science can be identified which call for  the development of specific techniques which account for the specificity of the data. Potential topics include e.g.  the investigation of structural interrelations among sets of social or economic actors, or data recorded over structured domains.

This seminar is designed to provide a thorough treatment of different aspects on network data including (1) the description, summary and visually representation present network data and (2) the formal econometric specification of the data through e.g. dyadic regression or exponential random graph models.  In addition, techniques for data scenarios where the domains are planar networks such as traffic networks will be investigated. Topics, literature and times will be fixed in an initial ZOOM meeting.

This seminar is restricted to at most 20 students and requires a personal registration via email to m.eckardt[at]hu-berlin.de.