Activity Number:
|
73
|
Type:
|
Contributed
|
Date/Time:
|
Sunday, July 31, 2016 : 4:00 PM to 5:50 PM
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
Abstract #321011
|
|
Title:
|
Asymptotic Relative Efficiency for Robust Estimation of the Mean of Contaminated Graphs Under a Low Rank Model
|
Author(s):
|
Runze Tang* and Minh Tang and Michael Ketcha and Carey Priebe and Joshua Vogelstein
|
Companies:
|
The Johns Hopkins University and The Johns Hopkins University and The Johns Hopkins University and The Johns Hopkins University and The Johns Hopkins University
|
Keywords:
|
Asymptotic Relative Efficiency ;
L-q Likelihood ;
Stochastic Blockmodel ;
Gross Error Model
|
Abstract:
|
To estimate the mean of a collection of weighted graphs under a low rank random graph model (e.g. Stochastic Blockmodel) when observing contaminated graphs, we propose an estimator which not only inherits robustness from element-wise robust estimators but also has small variance due to application of a rank-reduction procedure. Under appropriate conditions, we prove that our estimator outperforms standard estimators via asymptotic relative efficiency. We illustrate our theory and methods by Monte Carlo simulation studies and experimental results.
|
Authors who are presenting talks have a * after their name.