Significance testing in non-sparse high-dimensional linear models
From MaRDI portal
Publication:1616315
DOI10.1214/18-EJS1443zbMATH Open1416.62305arXiv1610.02122OpenAlexW2751732267WikidataQ129128194 ScholiaQ129128194MaRDI QIDQ1616315FDOQ1616315
Authors: Yinchu Zhu, Jelena Bradic 
Publication date: 1 November 2018
Published in: Electronic Journal of Statistics (Search for Journal in Brave)
Abstract: In high-dimensional linear models, the sparsity assumption is typically made, stating that most of the parameters are equal to zero. Under the sparsity assumption, estimation and, recently, inference have been well studied. However, in practice, sparsity assumption is not checkable and more importantly is often violated; a large number of covariates might be expected to be associated with the response, indicating that possibly all, rather than just a few, parameters are non-zero. A natural example is a genome-wide gene expression profiling, where all genes are believed to affect a common disease marker. We show that existing inferential methods are sensitive to the sparsity assumption, and may, in turn, result in the severe lack of control of Type-I error. In this article, we propose a new inferential method, named CorrT, which is robust to model misspecification such as heteroscedasticity and lack of sparsity. CorrT is shown to have Type I error approaching the nominal level for extit{any} models and Type II error approaching zero for sparse and many dense models. In fact, CorrT is also shown to be optimal in a variety of frameworks: sparse, non-sparse and hybrid models where sparse and dense signals are mixed. Numerical experiments show a favorable performance of the CorrT test compared to the state-of-the-art methods.
Full work available at URL: https://arxiv.org/abs/1610.02122
Recommendations
- Linear hypothesis testing in dense high-dimensional linear models
- Testability of high-dimensional linear models with nonsparse structures
- Statistical significance in high-dimensional linear models
- Testing regression coefficients in high-dimensional and sparse settings
- Testing a single regression coefficient in high dimensional linear models
Cites Work
- EigenPrism: inference for high dimensional signal-to-noise ratios
- Statistics for high-dimensional data. Methods, theory and applications.
- Detection boundary in sparse regression
- Confidence intervals for high-dimensional linear regression: minimax rates and adaptivity
- Simultaneous analysis of Lasso and Dantzig selector
- Confidence Intervals and Hypothesis Testing for High-Dimensional Regression
- Square-root lasso: pivotal recovery of sparse signals via conic programming
- Adapting to Unknown Smoothness via Wavelet Shrinkage
- Testing Statistical Hypotheses
- Title not available (Why is that?)
- Confidence Intervals for Low Dimensional Parameters in High Dimensional Linear Models
- A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity
- On asymptotically optimal confidence regions and tests for high-dimensional models
- Unified methods for censored longitudinal data and causality
- Hypothesis Testing in High-Dimensional Regression Under the Gaussian Random Design Model: Asymptotic Theory
- Scaled sparse linear regression
- Estimating Regression Models with Multiplicative Heteroscedasticity
- Estimation of Regression Coefficients When Some Regressors Are Not Always Observed
- The Asymptotic Variance of Semiparametric Estimators
- Asymptotic Statistics
- Semiparametric Regression for Repeated Outcomes with Nonignorable Nonresponse
- Inference on treatment effects after selection among high-dimensional controls
- Analysis of Semiparametric Regression Models for Repeated Outcomes in the Presence of Missing Data
- Concentration inequalities. A nonasymptotic theory of independence
- Minimax Rates of Estimation for High-Dimensional Linear Regression Over $\ell_q$-Balls
- A lava attack on the recovery of sums of dense and sparse signals
- Minimax risk over \(l_ p\)-balls for \(l_ q\)-error
- Minimax estimation of linear and quadratic functionals on sparsity classes
- Reconstruction From Anisotropic Random Measurements
- Title not available (Why is that?)
- Title not available (Why is that?)
- A general theory of hypothesis tests and confidence regions for sparse high dimensional models
- A Bernstein type inequality and moderate deviations for weakly dependent sequences
- Title not available (Why is that?)
- Semiparametric Efficiency in Multivariate Regression Models with Missing Data
- Semiparametric theory for causal mediation analysis: efficiency bounds, multiple robustness and sensitivity analysis
- Sharp adaptation for inverse problems with random noise
- Estimates for the distribution of sums and maxima of sums of random variables without the Cramér condition
- Doubly Robust and Efficient Estimators for Heteroscedastic Partially Linear Single-Index Models Allowing high Dimensional Covariates
- Pivotal Estimation in High-Dimensional Regression via Linear Programming
- Doubly robust learning for estimating individualized treatment with censored data
- Linear programming. Foundations and extensions
- Ridge regression and asymptotic minimax estimation over spheres of growing dimension
- Optimal adaptive estimation of linear functionals under sparsity
- Comments on: ``High-dimensional simultaneous inference with the bootstrap
- Debiasing the Lasso: optimal sample size for Gaussian designs
Cited In (24)
- Two-sample testing of high-dimensional linear regression coefficients via complementary sketching
- Relaxing the assumptions of knockoffs by conditioning
- A unified theory of confidence regions and testing for high-dimensional estimating equations
- Title not available (Why is that?)
- Comments on: ``High-dimensional simultaneous inference with the bootstrap
- A significance test for the lasso
- Projection-based Inference for High-dimensional Linear Models
- Testing a single regression coefficient in high dimensional linear models
- Linear hypothesis testing in dense high-dimensional linear models
- Optimal sparsity testing in linear regression model
- Testability of high-dimensional linear models with nonsparse structures
- CorrT
- In defense of the indefensible: a very naïve approach to high-dimensional inference
- Sign-based test for mean vector in high-dimensional and sparse settings
- Statistical significance in high-dimensional linear models
- Global testing under sparse alternatives: ANOVA, multiple comparisons and the higher criticism
- Double-estimation-friendly inference for high-dimensional misspecified models
- Distribution and correlation-free two-sample test of high-dimensional means
- Rejoinder on: ``High-dimensional simultaneous inference with the bootstrap
- Maximization of ESI. Jaynes principle in testing significant inputs of linear model
- A significance test for the elastic net and its asymptotic distribution with general predictors
- De-biasing the Lasso with degrees-of-freedom adjustment
- Permutation testing in high-dimensional linear models: an empirical investigation
- Estimation and Inference for High-Dimensional Generalized Linear Models with Knowledge Transfer
Uses Software
This page was built for publication: Significance testing in non-sparse high-dimensional linear models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1616315)
