Minimax Rate of Testing in Sparse Linear Regression

  title={Minimax Rate of Testing in Sparse Linear Regression},
  author={Alexandra Carpentier and Olivier Collier and Laetitia Comminges and A. Tsybakov and Yuhao Wang},
  journal={Automation and Remote Control},
  pages={1817 - 1834}
We consider the problem of testing the hypothesis that the parameter of linear regression model is 0 against an s-sparse alternative separated from 0 in the l2-distance. We show that, in Gaussian linear regression model with p < n, where p is the dimension of the parameter and n is the sample size, the non-asymptotic minimax rate of testing has the form (s/n)log(p/s)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage… 

High-dimensional asymptotics of likelihood ratio tests in the Gaussian sequence model under convex constraints

These characterizations show that the power behavior of the LRT is in general non-uniform with respect to the Euclidean metric, and illustrate the conservative nature of existing minimax optimality and sub-optimality results for the L RT.

Optimal sparsity testing in linear regression model

The minimax separation distances are pinpointed for the problem of sparsity testing in the high-dimensional linear regression model to test whether the number of non-zero components of the regression parameter $\theta^*$ is less than or equal to $k_0$.

Randomized tests for high-dimensional regression: A more efficient and powerful solution

This paper proposes a testing procedure that blends the classical $F$-test with a random projection step, and shows that the proposed test achieves sharp adaptive rates and is proved to be minimax optimal.

Testability of high-dimensional linear models with nonsparse structures.

It is shown that, in models with weak feature correlations, minimax lower bound can be attained by a test whose power has the n rate, regardless of the size of the model sparsity.


Under the null hypothesis, normal approximation holds for the log-likelihood ratio statistic for a general pair (μ0,K), in the high dimensional regime where the estimation error of the associated least squares estimator diverges in an appropriate sense.

Robust Testing in High-Dimensional Sparse Models

This work considers the problem of robustly testing the norm of a high-dimensional sparse signal vector under two different observation models and shows that the complexity of testing in these two settings significantly increases under robustness constraints.

Testing (Infinitely) Many Zero Restrictions

We present a max-test statistic for testing (possibly infinitely) many zero parameter restrictions in a general parametric regression framework. The test statistic is based on estimating the key

Estimation of the ℓ2-norm and testing in sparse linear regression with unknown variance

We consider the related problems of estimating the $l_2$-norm and the squared $l_2$-norm in sparse linear regression with unknown variance, as well as the problem of testing the hypothesis that the

Estimating the minimizer and the minimum value of a regression function under passive design

A new method for estimating the minimizer x ∗ and the minimum value f ∗ of a smooth and strongly convex regression function f from the observations contaminated by random noise and derive non- asymptotic upper bounds for the quadratic risk and optimization error of z n.

Near-Optimal Procedures for Model Discrimination with Non-Disclosure Properties

This work provides matching upper and lower bounds on the sample complexity as given by min{1/ ∆2,√r/∆} up to a constant factor; here ∆ is a measure of separation between P0 and P1 and r is the rank of the design covariance matrix.



Global testing under sparse alternatives: ANOVA, multiple comparisons and the higher criticism

Testing for the significance of a subset of regression coefficients in a linear model, a staple of statistical analysis, goes back at least to the work of Fisher who introduced the analysis of

Detection boundary in sparse regression

The main message is that, under some conditions, the detection boundary phenomenon that has been proved for the Gaussian sequence model, extends to high-dimensional linear regression.

Around the circular law

These expository notes are centered around the circular law theorem, which states that the empirical spectral distribution of a nxn random matrix with i.i.d. entries of variance 1/n tends to the

Introduction to the non-asymptotic analysis of random matrices

This is a tutorial on some basic non-asymptotic methods and concepts in random matrix theory, particularly for the problem of estimating covariance matrices in statistics and for validating probabilistic constructions of measurementMatrices in compressed sensing.

Nonparametric Goodness-of-Fit Testing Under Gaussian Models

This paper presents a meta-analyses of Gaussian Asymptotics for Power and Besov Norms and its applications to Minimax Distinguishability and high-Dimensional Signal Detection.

Minimax Estimation of Linear and Quadratic Functionals under Sparsity Constraints

  • Ann. Statist., 2017,
  • 2017

Non-asymptotic minimax rates of testing in signal detection

Let Y = (Yi)i∈I be a finite or countable sequence of independent Gaussian random variables of mean f = (fi)i∈I and common variance σ. For various sets F ⊂ `2(I), the aim of this paper is to describe

Higher criticism for detecting sparse heterogeneous mixtures

Higher criticism, or second-level significance testing, is a multiplecomparisons concept mentioned in passing by Tukey. It concerns a situation where there are many independent tests of significance

Minimax risks for sparse regressions: Ultra-high-dimensional phenomenons

This paper studies the minimax risks of estimation and testing over classes of $k$-sparse vectors $\theta$ and proves that even dimension reduction techniques cannot provide satisfying results in an ultra-high dimensional setting.