Two-Sample Hypothesis Testing for Subspace Equality in Network Data

2026-06-04

AI Summary

This study addresses the problem of testing whether two networks share the same underlying subspace structure—such as community organization—despite exhibiting markedly different edge probabilities. To this end, the authors propose a test statistic based on the Frobenius norm of the difference between subspace projection matrices and develop a unified testing framework applicable to both the stochastic block model and the mixed-membership stochastic block model. Leveraging spectral methods, matrix perturbation theory, and high-dimensional asymptotic analysis, they establish that, under the condition that the average expected degree grows at least logarithmically with network size, the standardized test statistic converges in distribution to a Gaussian limit. The efficacy and practical utility of the proposed method are corroborated through extensive simulations and an empirical application to U.S. airline flight data.

Abstract

In many settings one is often interested in determining whether two networks share some joint structural connectivity patterns such as communities. However, while communities may be shared across networks, edge probabilities may differ significantly. Therefore, in this paper we consider testing a general null hypothesis that two networks have the same underlying subspace, which in particular includes the setting that communities are the same for either stochastic blockmodels or mixed-membership stochastic blockmodels (even if edge probabilities are different). We propose a test statistic based on the Frobenius norm of the difference of the leading subspace projection matrices, and we prove that our test statistic, after appropriate centering and scaling, converges in distribution to a Gaussian random variable as long as the average expected degree grows at least logarithmically in the number of vertices. We then provide estimators for the asymptotic mean and variance and show consistency under a stronger signal condition, and we give the local power of our test when the networks are sufficiently dense. Our theoretical results are based on a limit theorem for the projection difference of empirical and true eigenvectors which can also be viewed as the one-sample version of our test statistic, and this result may be of independent interest. We demonstrate our results through numerical simulations and an application to US Flight data.

Problem

• Two-sample testing
• Subspace equality
• Network data
• Community structure
• Stochastic blockmodel

Innovation

• Two-sample hypothesis testing
• Subspace equality
• Network data
• Projection matrix
• Stochastic blockmodel