Inference robust to outliers with _1-norm penalization
From MaRDI portal
Publication:5140337
DOI10.1051/PS/2020014zbMATH Open1455.62065arXiv1906.01302OpenAlexW3014612614MaRDI QIDQ5140337FDOQ5140337
Authors: Jad Beyhum 
Publication date: 15 December 2020
Published in: ESAIM: Probability and Statistics (Search for Journal in Brave)
Abstract: This paper considers the problem of inference in a linear regression model with outliers where the number of outliers can grow with sample size but their proportion goes to 0. We apply the square-root lasso estimator penalizing the l1-norm of a random vector which is non-zero for outliers. We derive rates of convergence and asymptotic normality. Our estimator has the same asymptotic variance as the OLS estimator in the standard linear model. This enables to build tests and confidence sets in the usual and simple manner. The proposed procedure is also computationally advantageous as it amounts to solving a convex optimization program. Overall, the suggested approach constitutes a practical robust alternative to the ordinary least squares estimator.
Full work available at URL: https://arxiv.org/abs/1906.01302
Recommendations
- Robust censored regression with \(\ell_1\)-norm regularization
- Robust and sparse estimators for linear regression models
- Fully efficient robust estimation, outlier detection, and variable selection via penalized regression
- Sharp non-asymptotic performance bounds for \(\ell_1\) and Huber robust regression estimators
- Robust sparse regression with high-breakdown value
Linear regression; mixed models (62J05) Large deviations (60F10) Ridge regression; shrinkage estimators (Lasso) (62J07) Robustness and adaptive procedures (parametric inference) (62F35)
Cites Work
- Sparse least trimmed squares regression for analyzing high-dimensional large data sets
- Robust statistics. Theory and methods (with R)
- Square-root lasso: pivotal recovery of sparse signals via conic programming
- Robust Estimation of a Location Parameter
- Least squares after model selection in high-dimensional sparse models
- Scaled sparse linear regression
- Title not available (Why is that?)
- High-dimensional probability. An introduction with applications in data science
- Title not available (Why is that?)
- Robust regression through the Huber's criterion and adaptive lasso penalty
- Estimation of High Dimensional Mean Regression in the Absence of Symmetry and Light Tail Assumptions
- Regularization of case-specific parameters for robustness and efficiency
- Outlier detection using nonconvex penalized regression
- Title not available (Why is that?)
- Title not available (Why is that?)
- SOCP based variance free Dantzig selector with application to robust estimation
- Robust covariance and scatter matrix estimation under Huber's contamination model
Cited In (12)
- Robust and sparse estimators for linear regression models
- Gaining Outlier Resistance With Progressive Quantiles: Fast Algorithms and Theoretical Studies
- Fast approximate \(L_\infty\) minimization: speeding up robust regression
- Penalized trimmed squares and a modification of support vectors for unmasking outliers in linear regression
- Fully efficient robust estimation, outlier detection, and variable selection via penalized regression
- Penalized unsupervised learning with outliers
- High-dimensional inference robust to outliers with ℓ1-norm penalization
- Sharp non-asymptotic performance bounds for \(\ell_1\) and Huber robust regression estimators
- Efficient and robust estimation of regression and scale parameters, with outlier detection
- Robust estimation of skew-normal parameters with application to outlier labelling
- Robust censored regression with \(\ell_1\)-norm regularization
- Outlier detection using nonconvex penalized regression
Uses Software
This page was built for publication: Inference robust to outliers with \(\ell_1\)-norm penalization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5140337)
