A0479
Title: Generalized Bayesian inference for dynamic random dot product graphs
Authors: Joshua Loyal - Florida State University (United States) [presenting]
Abstract: The random dot product graph is a popular model for network data with extensions that accommodate dynamic (time-varying) networks. However, two significant deficiencies exist in the dynamic random dot product graph literature: (1) no coherent Bayesian way to update one's prior beliefs about the model parameters due to their complicated constraints, and (2) no approach to forecast future networks with meaningful uncertainty quantification. A generalized Bayesian framework is proposed that addresses these needs using a Gibbs posterior that represents a coherent updating of Bayesian beliefs based on a least-squares loss function. The consistency and contraction rate of this Gibbs posterior are established under commonly adopted Gaussian random walk priors. For estimation, a fast Gibbs sampler is developed with a time complexity that is linear in both the number of time points and observed edges in the dynamic network. Simulations and real data analyses show that the proposed methods in-sample and forecasting performance outperform that of competitors.
