Online Program Home
My Program

Abstract Details

Activity Number: 433
Type: Contributed
Date/Time: Tuesday, August 2, 2016 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistical Learning and Data Science
Abstract #318700 View Presentation
Title: Generalized Exponential Random Graph Models: Statistical Inference for Weighted Graphs
Author(s): James D. Wilson* and Shankar Bhamidi
Companies: University of San Francisco and The University of North Carolina at Chapel Hill
Keywords: random graphs ; networks ; exponential family ; correlation matrices

The exponential random graph model (ERGM) is a popular family of models for statistical inference with network data. However, a major limitation of the common formulation of the ERGM is that it can only be applied to networks with dichotomous edges. The generalized exponential random graph model (GERGM) is a flexible inferential tool to analyze networks with continuous-valued edge weights. The GERGM can be used to model a wide class of weighted networks and plays an important role in the understanding and analysis of relational data in fields such as genetics, neuroscience, social science, and finance. Additionally, the GERGM provides a means to model the generative process of correlation and partial correlation matrices. The topological modeling of correlation networks provides a new way of understanding populations of graphs that commonly arise in genetic and neuroscience studies. We describe the formulation of the GERGM and demonstrate its capabilities through two case studies including both social and functional brain networks. All applications will be performed using the newly released R package gergm.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association