Robust Estimation from Multiple Graphs under Gross Error Contamination
Methodology
2017-07-13 v1
Abstract
Estimation of graph parameters based on a collection of graphs is essential for a wide range of graph inference tasks. In practice, weighted graphs are generally observed with edge contamination. We consider a weighted latent position graph model contaminated via an edge weight gross error model and propose an estimation methodology based on robust Lq estimation followed by low-rank adjacency spectral decomposition. We demonstrate that, under appropriate conditions, our estimator both maintains Lq robustness and wins the bias-variance tradeoff by exploiting low-rank graph structure. We illustrate the improvement offered by our estimator via both simulations and a human connectome data experiment.
Cite
@article{arxiv.1707.03487,
title = {Robust Estimation from Multiple Graphs under Gross Error Contamination},
author = {Runze Tang and Minh Tang and Joshua T. Vogelstein and Carey E. Priebe},
journal= {arXiv preprint arXiv:1707.03487},
year = {2017}
}