Robust Regression and Outlier Detection

By Peter J. Rousseeuw, A. Leroy. Wiley Series in Probability and Statistics, 2005

8,700 citations recorded by the source corpus.

Subfields: Control and Systems Engineering, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/ae3f31c841c460b15a81bf51655f4c0e39cacc79

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

1. Michael J. Black
2. Jitendra Malik
3. Dimitris J. Bertsimas
4. Mathieu Salzmann
5. Mia Hubert
6. Guillermo Sapiro
7. Changliang Zou
8. Vincent Lepetit

In Statistics and Probability

1. Mia Hubert
2. Peter J. Rousseeuw
3. S. Van Aelst
4. Peter Filzmoser
5. Ilias Diakonikolas
6. Alistair Stewart
7. Changliang Zou
8. Jerry Zheng Li

Citations in the corpus

• White Rose Consortium ePrints Repository (neutral): “Unfortunately, it is susceptible to peeling away p + 1 points from the data distribution on each iteration and eroding the stock of normal points too rapidly (Rousseeuw and Leroy, 1996).”
• Robust Computation of Linear Models by Convex Relaxation (2012, background): “See the books [39,55,64] for an extensive discussion of other ways to combine residuals to obtain robust estimators.”
• Low-complexity fuzzy relational clustering algorithms for Web mining (2001, methodology): “the Least Trimmed Squares idea [42], [28], we use the membership function in (2).”
• Robust Estimators are Hard to Compute (2006, background): “The mentioned estimators match the definition of the subset estima- tor: • LMS, LQS, LTS and LTA: If a point is located on a hyperplane, its residual is zero.”
• Best-Practice Recommendations for Defining, Identifying, and Handling Outliers (2013, neutral): “Email: haguinis@indiana.edu at University of Exeter on May 2, 2014orm.sagepub.comDownloaded from Jackman, 1990; Grubbs, 1969; Hunter & Schmidt, 2004; Rousseeuw & Leroy, 2003).”
• The Deepest Regression Method (2002, background): “;j = 1,...,m}, where the robust distance (Rousseeuw and Leroy [ 21 ]) of ˆ �”
• Robust supervised classification with mixture models: Learning from data with uncertain labels (2009, methodology): “Maximum likelihood estimators of the mixture model parameters are replaced by the corresponding S-estimators (see Rousseeuw and Leroy [25] for a general account on robust estimation) but the authors only observed a slight reduction of the average probability of misclassification.”
• Modelling the innovation value chain (2008, methodology): “To counter the bias introduced by potential outliers we here adopt robust regression approaches to the estimation of the augmented production function (Rousseeuw and Leroy, 1987; Koenker and Bassett, 1978).”
• Comprehensive definitions of breakdown points for independent and dependent observations (2003, background): “See also the arguments raised in Stromberg and Ruppert (1992) and Sakata and White (1995). Following Sakata and White (1995), therefore, we introduce a measure of badness R.”
• Monitoring robust regression (2014, methodology): “For example, Rousseeuw and Leroy (1987), p. 111, present comparative index plots of least squares (LS) and least median of squares (LMS) residuals.”
• Aid and growth in Sub-Saharan Africa: Accounting for transmission mechanisms (2005, methodology): “We have here chosen an alternative method – robust regression (Rousseeuw and Leroy, 1987), detailed in Appendix B. Robust regression, which retain all observations but attaches lower weight to outliers, is used throughout this paper.”