SteinGen: Generating Fidelitous and Diverse Graph Samples

arXiv:2403.18578v1 Announce Type: cross
Abstract: Generating graphs that preserve characteristic structures while promoting sample diversity can be challenging, especially when the number of graph observations is small. Here, we tackle the problem of graph generation from only one observed graph. The classical approach of graph generation from parametric models relies on the estimation of parameters, which can be inconsistent or expensive to compute due to intractable normalisation constants. Generative modelling based on machine learning techniques to generate high-quality graph samples …

abstract arxiv cs.lg diverse diversity graph graphs parametric sample samples small stat.ml type