Graphical Assistant Grouped Network Autoregression Model: a Bayesian Nonparametric Recourse

TL;DR

This paper introduces a Bayesian nonparametric grouped network autoregression model that leverages a graphical Chinese restaurant process for improved inference on time series data over latent graphs, with applications to social network analysis.

Contribution

It develops a novel Bayesian grouped network autoregression model incorporating a graphical Chinese restaurant process for better group estimation and inference.

Findings

Effective in estimating group structures in simulated data

Demonstrates improved inference performance over existing methods

Successfully applied to real social network datasets

Abstract

Vector autoregression model is ubiquitous in classical time series data analysis. With the rapid advance of social network sites, time series data over latent graph is becoming increasingly popular. In this paper, we develop a novel Bayesian grouped network autoregression model to simultaneously estimate group information (number of groups and group configurations) and group-wise parameters. Specifically, a graphically assisted Chinese restaurant process is incorporated under framework of the network autoregression model to improve the statistical inference performance. An efficient Markov chain Monte Carlo sampling algorithm is used to sample from the posterior distribution. Extensive studies are conducted to evaluate the finite sample performance of our proposed methodology. Additionally, we analyze two real datasets as illustrations of the effectiveness of our approach.