Robust subset selection

Abstract

The best subset selection (or "best subsets") estimator is a classic tool for sparse regression, and developments in mathematical optimization over the past decade have made it more computationally tractable than ever. Notwithstanding its desirable statistical properties, the best subsets estimator is susceptible to outliers and can break down in the presence of a single contaminated data point. To address this issue, a robust adaption of best subsets is proposed that is highly resistant to contamination in both the response and the predictors. The adapted estimator generalizes the notion of subset selection to both predictors and observations, thereby achieving robustness in addition to sparsity. This procedure, referred to as "robust subset selection" (or "robust subsets"), is defined by a combinatorial optimization problem for which modern discrete optimization methods are applied. The robustness of the estimator in terms of the finite-sample breakdown point of its objective value is formally established. In support of this result, experiments on synthetic and real data are reported that demonstrate the superiority of robust subsets over best subsets in the presence of contamination. Importantly, robust subsets fares competitively across several metrics compared with popular robust adaptions of continuous shrinkage estimators.

arXiv commentTo appear in Computational Statistics & Data Analysis

References in corpus (20)

• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Rejoinder to "Least angle regression" by Efron et al
• Least Angle Regression
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Nearly unbiased variable selection under minimax concave penalty
• Pathwise coordinate optimization
• On the conditions used to prove oracle results for the Lasso
• Piecewise linear regularized solution paths
• Sparse least trimmed squares regression for analyzing high-dimensional large data sets
• Aggregation for Gaussian regression
• Robust Regression via Hard Thresholding
• Sparse Regression: Scalable algorithms and empirical performance
• Least quantile regression via modern optimization
• MIP-BOOST: Efficient and Effective $L_0$ Feature Selection for Linear Regression