|
Activity Number:
|
495
- Statistical Methods for Networks
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
|
Abstract #312498
|
|
|
Title:
|
Consistent Nonparametric Hypothesis Testing for Low Rank Random Graphs with Negative or Repeated Eigenvalues
|
|
Author(s):
|
Joshua Agterberg* and Minh Tang and Carey Priebe and Mao Hong
|
|
Companies:
|
Johns Hopkins University and NC State University and Johns Hopkins University and Johns Hopkins University
|
|
Keywords:
|
Networks;
Random Graphs;
Hypothesis Testing;
Nonparametric;
Kernel Methods
|
|
Abstract:
|
Motivated in part by the indefinite homogeneous balanced stochastic block model, we propose a nonparametric test for testing equality of distributions for random graphs whose edge probability matrices may have repeated or negative eigenvalues. Our proposed methodology involves using a kernel-based function of the optimally rotated spectral embeddings of the graphs, where the rotation is determined as the solution to an associated optimal transport problem. We show the consistency of our proposed estimator and demonstrate its effectiveness on real and simulated data.
|
Authors who are presenting talks have a * after their name.
