Testability of high-dimensional linear models with non-sparse structures

Abstract

Understanding statistical inference under possibly non-sparse high-dimensional models has gained much interest recently. For a given component of the regression coefficient, we show that the difficulty of the problem depends on the sparsity of the corresponding row of the precision matrix of the covariates, not the sparsity of the regression coefficients. We develop new concepts of uniform and essentially uniform non-testability that allow the study of limitations of tests across a broad set of alternatives. Uniform non-testability identifies a collection of alternatives such that the power of any test, against any alternative in the group, is asymptotically at most equal to the nominal size. Implications of the new constructions include new minimax testability results that, in sharp contrast to the current results, do not depend on the sparsity of the regression parameters. We identify new tradeoffs between testability and feature correlation. In particular, we show that, in models with weak feature correlations, minimax lower bound can be attained by a test whose power has the $\sqrt{n}$ rate, regardless of the size of the model sparsity.

References in corpus (7)

• On adaptive inference and confidence bands
• High-Dimensional Graphical Model Selection Using ${\ell }_1$-Regularized Logistic Regression
• Adaptive nonparametric confidence sets
• Adaptive confidence balls
• Linear Hypothesis Testing in Dense High-Dimensional Linear Models
• Goodness of fit tests for high-dimensional linear models
• Significance testing in non-sparse high-dimensional linear models

Cited by in corpus (11)

• Significance testing in non-sparse high-dimensional linear models
• De-Biasing The Lasso With Degrees-of-Freedom Adjustment
• CAD: Debiasing the Lasso with inaccurate covariate model
• Detection and inference of changes in high-dimensional linear regression with non-sparse structures
• How well can we learn large factor models without assuming strong factors?
• Chi-square and normal inference in high-dimensional multi-task regression
• A New Perspective on Debiasing Linear Regressions
• Semi-supervised linear regression: enhancing efficiency and robustness in high dimensions
• Honest confidence sets for high-dimensional regression by projection and shrinkage
• Near-Optimal Procedures for Model Discrimination with Non-Disclosure Properties
• Comments on Leo Breiman's paper 'Statistical Modeling: The Two Cultures' (Statistical Science, 2001, 16(3), 199-231)