A Nonparametric Bayesian Model for Sparse Dynamic Multigraphs
arXiv:1910.05098
Abstract
As the availability and importance of temporal interaction data--such as email communication--increases, it becomes increasingly important to understand the underlying structure that underpins these interactions. Often these interactions form a multigraph, where we might have multiple interactions between two entities. Such multigraphs tend to be sparse yet structured, and their distribution often evolves over time. Existing statistical models with interpretable parameters can capture some, but not all, of these properties. We propose a dynamic nonparametric model for interaction multigraphs that combines the sparsity of edge-exchangeable multigraphs with dynamic clustering patterns that tend to reinforce recent behavioral patterns. We show that our method yields improved held-out likelihood over stationary variants, and impressive predictive performance against a range of state-of-the-art dynamic graph models.
References in corpus (14)
- Stochastic blockmodels and community structure in networks
- Dynamic stochastic blockmodels for time-evolving social networks
- Sparse graphs using exchangeable random measures
- Nonparametric Bayes dynamic modeling of relational data
- Infinite Edge Partition Models for Overlapping Community Detection and Link Prediction
- Generalized Polya Urn for Time-varying Dirichlet Process Mixtures
- The Class of Random Graphs Arising from Exchangeable Random Measures
- Private Graphon Estimation for Sparse Graphs
- Nonparametric Bayesian Factor Analysis for Dynamic Count Matrices
- Edge-exchangeable graphs and sparsity
- Sampling perspectives on sparse exchangeable graphs
- A Poisson Gamma Probabilistic Model for Latent Node-group Memberships in Dynamic Networks
- A Dynamic Edge Exchangeable Model for Sparse Temporal Networks
- A Bayesian model for sparse graphs with flexible degree distribution and overlapping community structure