On the conditional distributions of low-dimensional projections from high-dimensional data

DOI10.1214/12-AOS1081MaRDI QIDQ355082zbMATH OpenOpenAlexFDO

Authors Hannes Leeb

Publication date 24 July 2013

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1304.5943, https://projecteuclid.org/euclid.aos/1366138198

zbMATH Keywords

dimension reductionregressionhigh-dimensional modelssmall sample size

Mathematics Subject Classification ID

Multivariate analysis (62H99) Asymptotic distribution theory in statistics (62E20) Estimation in multivariate analysis (62H12)

Abstract: We study the conditional distribution of low-dimensional projections from high-dimensional data, where the conditioning is on other low-dimensional projections. To fix ideas, consider a random d-vector Z that has a Lebesgue density and that is standardized so that $mathbbEZ=0$ and $mathbbEZ{Z}^{\prime }=I_d$. Moreover, consider two projections defined by unit-vectors $alpha$ and $$, namely a response $y=alph{a}^{\prime }Z$ and an explanatory variable $$. It has long been known that the conditional mean of y given x is approximately linear in x$undersomeregularityconditions;cf.HallandLi[Ann.Statist.21(1993)867-889].However,acorrespondingresultfortheconditionalvariancehasnotbeenavailablesofar.Wehereshowthattheconditionalvarianceofygivenxisapproximatelyconstantinx(again,undersomeregularityconditions).Theseresultsholduniformlyin$alpha$andformost$�eta$'s, provided only that the dimension of Z is large. In that sense, we see that most linear submodels of a high-dimensional overall model are approximately correct. Our findings provide new insights in a variety of modeling scenarios. We discuss several examples, including sliced inverse regression, sliced average variance estimation, generalized linear models under potential link violation, and sparse linear modeling.

Recommendations

• On almost linearity of low dimensional projections from high dimensional data
• Approximation of projections of random vectors
• On conditional moments of high-dimensional random vectors given lower-dimensional projections
• On Sliced Inverse Regression With High-Dimensional Covariates
• On low-dimensional projections of high-dimensional distributions

Cites work

• A characterization of spherical distributions
• Asymptotics of graphical projection pursuit
• Coordinate-independent sparse sufficient dimension reduction and variable selection
• Dimension reduction for non-elliptically distributed predictors: second-order methods
• Nonlinear confounding in high-dimensional regression
• On Directional Regression for Dimension Reduction
• On Principal Hessian Directions for Data Visualization and Dimension Reduction: Another Application of Stein's Lemma
• On almost linearity of low dimensional projections from high dimensional data
• On low-dimensional projections of high-dimensional distributions
• On the conditional distributions of low-dimensional projections from high-dimensional data
• Principal support vector machines for linear and nonlinear sufficient dimension reduction
• Real Analysis and Probability
• Regression analysis under link violation
• Sliced Inverse Regression for Dimension Reduction
• Sufficient dimension reduction in regressions with categorical predictors
• The Large-Sample Power of Tests Based on Permutations of Observations
• The normal distribution. Characterizations with applications

Cited in

(7)

• An Inverse-regression Method of Dependent Variable Transformation for Dimension Reduction with Non-linear Confounding
• Projective inference in high-dimensional problems: prediction and feature selection
• On almost linearity of low dimensional projections from high dimensional data
• STATISTICAL INFERENCE WITH F-STATISTICS WHEN FITTING SIMPLE MODELS TO HIGH-DIMENSIONAL DATA
• On the conditional distributions of low-dimensional projections from high-dimensional data
• On conditional moments of high-dimensional random vectors given lower-dimensional projections
• A Random Projection Approach to Hypothesis Tests in High-Dimensional Single-Index Models

This page was built for publication: On the conditional distributions of low-dimensional projections from high-dimensional data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q355082)