A fast algorithm for robust regression with penalised trimmed squares

DOI10.1007/S00180-010-0196-2zbMATH Open1416.62434arXiv0901.0876OpenAlexW3103377816MaRDI QIDQ650724FDOQ650724

Authors: Leonidas Pitsoulis, George Zioutas 

Publication date: 26 November 2011

Published in: Computational Statistics (Search for Journal in Brave)

Abstract: The presence of groups containing high leverage outliers makes linear regression a difficult problem due to the masking effect. The available high breakdown estimators based on Least Trimmed Squares often do not succeed in detecting masked high leverage outliers in finite samples. An alternative to the LTS estimator, called Penalised Trimmed Squares (PTS) estimator, was introduced by the authors in cite{ZiouAv:05,ZiAvPi:07} and it appears to be less sensitive to the masking problem. This estimator is defined by a Quadratic Mixed Integer Programming (QMIP) problem, where in the objective function a penalty cost for each observation is included which serves as an upper bound on the residual error for any feasible regression line. Since the PTS does not require presetting the number of outliers to delete from the data set, it has better efficiency with respect to other estimators. However, due to the high computational complexity of the resulting QMIP problem, exact solutions for moderately large regression problems is infeasible. In this paper we further establish the theoretical properties of the PTS estimator, such as high breakdown and efficiency, and propose an approximate algorithm called Fast-PTS to compute the PTS estimator for large data sets efficiently. Extensive computational experiments on sets of benchmark instances with varying degrees of outlier contamination, indicate that the proposed algorithm performs well in identifying groups of high leverage outliers in reasonable computational time.

Full work available at URL: https://arxiv.org/abs/0901.0876

Recommendations

• Penalized trimmed squares and a modification of support vectors for unmasking outliers in linear regression
• New algorithms for computing the least trimmed squares regression estimator
• On the least trimmed squares estimator
• The feasible solution algorithm for least trimmed squares regression
• Quadratic mixed integer programming and support vectors for deleting outliers in robust regression

zbMATH Keywords

robust regressionleast trimmed squaresoutliers detectionquadratic mixed integer programming

Mathematics Subject Classification ID

Diagnostics, and linear inference and regression (62J20) Robustness and adaptive procedures (parametric inference) (62F35) Mixed integer programming (90C11)

Cites Work

• BACON: blocked adaptive computationally efficient outlier nominators.
• Title not available (Why is that?)
• Least Median of Squares Regression
• Title not available (Why is that?)
• High breakdown-point and high efficiency robust estimates for regression
• A class of robust and fully efficient regression estimators
• Fast Very Robust Methods for the Detection of Multiple Outliers
• Improved feasible solution algorithms for high breakdown estimation.
• High Breakdown-Point Estimates of Regression by Means of the Minimization of an Efficient Scale
• A one-step robust estimator for regression based on the weighted likelihood reweighting scheme
• Deleting outliers in robust regression with mixed integer programming
• Greedy randomized adaptive search procedures
• Least trimmed squares regression, least median squares regression, and mathematical program\-ming
• Title not available (Why is that?)
• New algorithms for computing the least trimmed squares regression estimator
• A Bounded Influence, High Breakdown, Efficient Regression Estimator
• The feasible solution algorithm for least trimmed squares regression
• A Comparison of Multiple Outlier Detection Methods for Regression Data
• Penalized trimmed squares and a modification of support vectors for unmasking outliers in linear regression
• Detecting Outliers. II. Supplementing the Direct Analysis of Residuals
• A clustering algorithm for identifying multiple outliers in linear regression.
• A Re-Weighted Least Squares Method for Robust Regression Estimation
• A Fast Procedure for Outlier Diagnostics in Large Regression Problems
• An algorithm for computing exact least-trimmed squares estimate of simple linear regression with constraints

Cited In (5)

• Optimization techniques for robust multivariate location and scatter estimation
• Penalized trimmed squares and a modification of support vectors for unmasking outliers in linear regression
• Leveraged least trimmed absolute deviations
• A comparison of some quick algorithms for robust regression
• A robust regression based on weighted LSSVM and penalized trimmed squares

Uses Software

• mvBACON

This page was built for publication: A fast algorithm for robust regression with penalised trimmed squares

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q650724)