Towards quantitative methods to assess network generative models
This addresses the challenge of quantitatively assessing generative models for network synthesis, which is incremental as it adapts existing classifier methods to a new evaluation context.
The paper tackles the problem of evaluating graph generative models by proposing an approach that uses graph classifiers to measure how well synthesized graphs mimic real networks, with the inability to distinguish them serving as a performance metric.
Assessing generative models is not an easy task. Generative models should synthesize graphs which are not replicates of real networks but show topological features similar to real graphs. We introduce an approach for assessing graph generative models using graph classifiers. The inability of an established graph classifier for distinguishing real and synthesized graphs could be considered as a performance measurement for graph generators.