Open Access
Translator Disclaimer
October 2013 High-dimensional influence measure
Junlong Zhao, Chenlei Leng, Lexin Li, Hansheng Wang
Ann. Statist. 41(5): 2639-2667 (October 2013). DOI: 10.1214/13-AOS1165


Influence diagnosis is important since presence of influential observations could lead to distorted analysis and misleading interpretations. For high-dimensional data, it is particularly so, as the increased dimensionality and complexity may amplify both the chance of an observation being influential, and its potential impact on the analysis. In this article, we propose a novel high-dimensional influence measure for regressions with the number of predictors far exceeding the sample size. Our proposal can be viewed as a high-dimensional counterpart to the classical Cook’s distance. However, whereas the Cook’s distance quantifies the individual observation’s influence on the least squares regression coefficient estimate, our new diagnosis measure captures the influence on the marginal correlations, which in turn exerts serious influence on downstream analysis including coefficient estimation, variable selection and screening. Moreover, we establish the asymptotic distribution of the proposed influence measure by letting the predictor dimension go to infinity. Availability of this asymptotic distribution leads to a principled rule to determine the critical value for influential observation detection. Both simulations and real data analysis demonstrate usefulness of the new influence diagnosis measure.


Download Citation

Junlong Zhao. Chenlei Leng. Lexin Li. Hansheng Wang. "High-dimensional influence measure." Ann. Statist. 41 (5) 2639 - 2667, October 2013.


Published: October 2013
First available in Project Euclid: 19 November 2013

zbMATH: 1360.62411
MathSciNet: MR3161440
Digital Object Identifier: 10.1214/13-AOS1165

Primary: 62J20
Secondary: 62E20

Keywords: Cook’s distance , high-dimensional diagnosis , influential observation , Lasso , marginal correlations , variable screening

Rights: Copyright © 2013 Institute of Mathematical Statistics


Vol.41 • No. 5 • October 2013
Back to Top