|
Activity Number:
|
101
- Network Analytics in the Era of Big Data
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, July 30, 2018 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Marketing
|
|
Abstract #326901
|
Presentation
|
|
Title:
|
Estimation of Change Point in Temporally Evolving Networks
|
|
Author(s):
|
Moulinath Banerjee* and George Michailidis and Monika Bhatacharjee
|
|
Companies:
|
University of Michigan and University of Florida and University of Florida
|
|
Keywords:
|
change point;
clustering;
stochastic block models
|
|
Abstract:
|
We consider a dynamic stochastic block model for a temporally evolving network, that exhibits a single change point at some point during the observation time period. An easily implementable algorithm based on a maximum pseudo-likelihood method coupled with spectral clustering is proposed for estimating the change point, as well as edge-probability matrices and clustering functions before and after the identified change point. The convergence rate and asymptotic distribution for these estimators are discussed and compared with other existing works in the literature. The performance of the method is illustrated on synthetic data sets and an application of voting records of the US Congress.
|
Authors who are presenting talks have a * after their name.
