Open Access
June 2010 Empirical stationary correlations for semi-supervised learning on graphs
Ya Xu, Justin S. Dyer, Art B. Owen
Ann. Appl. Stat. 4(2): 589-614 (June 2010). DOI: 10.1214/09-AOAS293

Abstract

In semi-supervised learning on graphs, response variables observed at one node are used to estimate missing values at other nodes. The methods exploit correlations between nearby nodes in the graph. In this paper we prove that many such proposals are equivalent to kriging predictors based on a fixed covariance matrix driven by the link structure of the graph. We then propose a data-driven estimator of the correlation structure that exploits patterns among the observed response values. By incorporating even a small fraction of observed covariation into the predictions, we are able to obtain much improved prediction on two graph data sets.

Citation

Download Citation

Ya Xu. Justin S. Dyer. Art B. Owen. "Empirical stationary correlations for semi-supervised learning on graphs." Ann. Appl. Stat. 4 (2) 589 - 614, June 2010. https://doi.org/10.1214/09-AOAS293

Information

Published: June 2010
First available in Project Euclid: 3 August 2010

zbMATH: 1194.62083
MathSciNet: MR2758641
Digital Object Identifier: 10.1214/09-AOAS293

Keywords: graph Laplacian , kriging , PageRank , Random walk

Rights: Copyright © 2010 Institute of Mathematical Statistics

Vol.4 • No. 2 • June 2010
Back to Top