Title: Robust continuum regression
Authors: Serneels, S
Filzmoser, P
Croux, Christophe
Van Espen, P
Issue Date: 2004
Publisher: K.U.Leuven - Departement toegepaste economische wetenschappen
Series Title: DTEW Research Report 0456 pages:1-28
Abstract: Several applications of continuum regression to non-contaminated data have shown that a significant improvement in predictive power can be obtained compared to the three standard techniques which it encompasses (Ordinary least Squares, Principal Component Regression and Partial Least Squares). For contaminated data continuum regression may yield aberrant estimates due to its non-robustness with respect to outliers. Also for data originating from a distribution which significantly differs from the normal distribution, continuum regression may yield very inefficient estimates. In the current paper, robust continuum regression (RCR) is proposed. To construct the estimator, an algorithm based on projection pursuit is proposed. The robustness and good efficiency properties of RCR are shown by means of a simulation study. An application to an X-ray fluorescence analysis of hydrometallurgical samples illustrates the method's applicability in practice.
Publication status: published
KU Leuven publication type: IR
Appears in Collections:Research Center for Operations Research and Business Statistics (ORSTAT), Leuven

Files in This Item:
File Status SizeFormat
OR_0456.pdf Published 569KbAdobe PDFView/Open


All items in Lirias are protected by copyright, with all rights reserved.