Network cross-validation by edge sampling

Abstract

While many statistical models and methods are now available for network analysis, resampling network data remains a challenging problem. Cross-validation is a useful general tool for model selection and parameter tuning, but is not directly applicable to networks since splitting network nodes into groups requires deleting edges and destroys some of the network structure. Here we propose a new network resampling strategy based on splitting node pairs rather than nodes applicable to cross-validation for a wide range of network model selection tasks. We provide a theoretical justification for our method in a general setting and examples of how our method can be used in specific network model selection and parameter tuning tasks. Numerical results on simulated networks and on a citation network of statisticians show that this cross-validation approach works well for model selection.

References in corpus (13)

• Stochastic blockmodels and community structure in networks
• Structure and inference in annotated networks
• Graph limits and exchangeable random graphs
• Modeling homophily and stochastic equivalence in symmetric relational data
• Consistency of cross validation for comparing regression procedures
• Statistical inference on random dot product graphs: a survey
• Optimal Estimation and Completion of Matrices with Biclustering Structures
• Entrywise Eigenvector Analysis of Random Matrices with Low Expected Rank
• Stochastic Block Model and Community Detection in the Sparse Graphs: A spectral algorithm with optimal rate of recovery
• Estimating the number of communities in networks by spectral methods
• Estimating network edge probabilities by neighborhood smoothing
• 1-Bit Matrix Completion under Exact Low-Rank Constraint
• Cross-Validation with Confidence

Cited by in corpus (7)

• Estimating the number of communities in networks by spectral methods
• Bootstrapping Networks with Latent Space Structure
• Determining the Number of Communities in Degree-corrected Stochastic Block Models
• Next Waves in Veridical Network Embedding
• Selection of Exponential-Family Random Graph Models via Held-Out Predictive Evaluation (HOPE)
• Bayesian Model Selection on Random Networks
• Network estimation via graphon with node features