Which bridge estimator is optimal for variable selection?

Abstract

We study the problem of variable selection for linear models under the high-dimensional asymptotic setting, where the number of observations $n$ grows at the same rate as the number of predictors $p$. We consider two-stage variable selection techniques (TVS) in which the first stage uses bridge estimators to obtain an estimate of the regression coefficients, and the second stage simply thresholds this estimate to select the "important" predictors. The asymptotic false discovery proportion (AFDP) and true positive proportion (ATPP) of these TVS are evaluated. We prove that for a fixed ATPP, in order to obtain a smaller AFDP, one should pick a bridge estimator with smaller asymptotic mean square error in the first stage of TVS. Based on such principled discovery, we present a sharp comparison of different TVS, via an in-depth investigation of the estimation properties of bridge estimators. Rather than "order-wise" error bounds with loose constants, our analysis focuses on precise error characterization. Various interesting signal-to-noise ratio and sparsity settings are studied. Our results offer new and thorough insights into high-dimensional variable selection. For instance, we prove that a TVS with Ridge in its first stage outperforms TVS with other bridge estimators in large noise settings; two-stage LASSO becomes inferior when the signal is rare and weak. As a by-product, we show that two-stage methods outperform some standard variable selection techniques, such as LASSO and Sure Independence Screening, under certain conditions.

84 pages, 11 figures

References in corpus (5)

• Asymptotic properties of bridge estimators in sparse high-dimensional regression models
• Extended Comparisons of Best Subset Selection, Forward Stepwise Selection, and the Lasso
• Higher Criticism for Large-Scale Inference, Especially for Rare and Weak Effects
• High-dimensionality effects in the Markowitz problem and other quadratic programs with linear constraints: Risk underestimation
• High-dimensional variable selection via tilting

Cited by in corpus (7)

• Approximate Message Passing for Amplitude Based Optimization
• Approximate Leave-One-Out for High-Dimensional Non-Differentiable Learning Problems
• Out-of-sample error estimate for robust M-estimators with convex penalty
• Matrix Completion with Nonconvex Regularization: Spectral Operators and Scalable Algorithms
• Estimation of Switched Markov Polynomial NARX models
• Fully Bayesian Penalized Regression with a Generalized Bridge Prior
• Low noise sensitivity analysis of Lq-minimization in oversampled systems