|
Activity Number:
|
557
- New Directions in Bayesian Methods for Longitudinal and Graph Data
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 11, 2022 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
|
Abstract #323049
|
|
|
Title:
|
Inference of Random Dot Product Graphs with a Surrogate Likelihood Function
|
|
Author(s):
|
Dingbo Wu* and Fangzheng Xie
|
|
Companies:
|
Indiana University and Indiana University
|
|
Keywords:
|
random dot product graph;
surrogate likelihood;
generalized Bayesian inference;
stochastic gradient descent;
Markov chain Monte Carlo
|
|
Abstract:
|
We propose a surrogate likelihood function for random dot product graphs. This surrogate likelihood makes use of both the low-rank structure of the expectation of adjacency matrix and the Bernoulli likelihood information of the sampling model. Based on the surrogate likelihood, we implement two inference procedures: maximum surrogate likelihood estimation and generalized Bayesian inference with surrogate likelihood. We establish the asymptotic properties of the proposed inference procedures. We study the finite sample performance of the statistics from the inference procedures with some simulation examples and the analysis of an empirical Wikipedia graph dataset.
|
Authors who are presenting talks have a * after their name.
