Research interest

Graph generative models

Learning distributions over graphs, from molecules to networks, and checking whether samples look like the data.

Reading and experiments · Aalborg University

Draft. This page is mostly placeholder: add what you're working on here.

Does a sample look like the data? Degree distribution of a two-block graph (grey) against samples from a fitted SBM (blue) and from an Erdős–Rényi model with the same density (dashed). Only the SBM reproduces both peaks.

The question

A generative model of graphs is only as good as the statistics its samples get right. Degree distributions, clustering coefficients and spectra are the usual checks, compared with measures such as maximum mean discrepancy.

What I'm doing

[Your current reading, experiments or planned project.]