Robust rank correlation based screening

Abstract

Independence screening is a variable selection method that uses a ranking criterion to select significant variables, particularly for statistical models with nonpolynomial dimensionality or "large p, small n" paradigms when p can be as large as an exponential of the sample size n. In this paper we propose a robust rank correlation screening (RRCS) method to deal with ultra-high dimensional data. The new procedure is based on the Kendall τcorrelation coefficient between response and predictor variables rather than the Pearson correlation of existing methods. The new method has four desirable features compared with existing independence screening methods. First, the sure independence screening property can hold only under the existence of a second order moment of predictor variables, rather than exponential tails or alikeness, even when the number of predictor variables grows as fast as exponentially of the sample size. Second, it can be used to deal with semiparametric models such as transformation regression models and single-index models under monotonic constraint to the link function without involving nonparametric estimation even when there are nonparametric functions in the models. Third, the procedure can be largely used against outliers and influence points in the observations. Last, the use of indicator functions in rank correlation screening greatly simplifies the theoretical derivation due to the boundedness of the resulting statistics, compared with previous studies on variable screening. Simulations are carried out for comparisons with existing methods and a real data example is analyzed.

Published in at http://dx.doi.org/10.1214/12-AOS1024 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org). arXiv admin note: text overlap with arXiv:0903.5255

References in corpus (6)

• One-step sparse estimates in nonconcave penalized likelihood models
• Composite quantile regression and the oracle Model Selection Theory
• Asymptotic properties of bridge estimators in sparse high-dimensional regression models
• High-dimensional generalized linear models and the lasso
• Statistical Challenges with High Dimensionality: Feature Selection in Knowledge Discovery
• A unified approach to model selection and sparse recovery using regularized least squares

Cited by in corpus (35)

• Challenges of Big Data Analysis
• The fused Kolmogorov filter: A nonparametric model-free screening method
• Marginal empirical likelihood and sure independence feature screening
• Least Squares Approximation for a Distributed System
• Nonparametric independence screening and structure identification for ultra-high dimensional longitudinal data
• An Atomistic Machine Learning Package for Surface Science and Catalysis
• Local independence feature screening for nonparametric and semiparametric models by marginal empirical likelihood
• Cramér-type moderate deviations for Studentized two-sample $U$-statistics with applications
• Doubly Robust Sure Screening for Elliptical Copula Regression Model
• Variable Screening for High Dimensional Time Series
• Distribution-free and Model-free Multivariate Feature Screening via Multivariate Rank Distance Correlation
• Nonparametric Independence Screening via Favored Smoothing Bandwidth
• ExSIS: Extended Sure Independence Screening for Ultrahigh-dimensional Linear Models
• The Kendall Interaction Filter for Variable Interaction Screening in Ultra High Dimensional Classification Problems
• Nonparametric Screening under Conditional Strictly Convex Loss for Ultrahigh Dimensional Sparse Data
• Sure screening for estimating equations in ultra-high dimensions
• On the consistency theory of high dimensional variable screening
• Comparisons of penalized least squares methods by simulations
• Covariate Assisted Variable Ranking
• Copula-based Partial Correlation Screening: a Joint and Robust Approach
• Independent screening for single-index hazard rate models with ultra-high dimensional features
• A Simplified Algorithm for Identifying Abnormal Changes in Dynamic Networks
• Model-Free Conditional Feature Screening with Exposure Variables
• Robust variable screening for regression using factor profiling
• Greedy Forward Regression for Variable Screening
• Optimal Estimation of Slope Vector in High-dimensional Linear Transformation Model
• Covariance-Insured Screening
• Ultra High Dimensional Change Point Detection
• Trace Pursuit: A General Framework for Model-Free Variable Selection
• Better Solution Principle: A Facet of Concordance between Optimization and Statistics
• A General Framework of Nonparametric Feature Selection in High-Dimensional Data
• Conditional variable screening via ordinary least squares projection
• A sure independence screening procedure for ultra-high dimensional partially linear additive models
• Fused Mean-variance Filter for Feature Screening
• High-dimensional Ordinary Least-squares Projection for Screening Variables