|
Activity Number:
|
545
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #310258 |
|
Title:
|
A New Robust Partial Least Squares Regression Method (RoPLS) and Its Robustness
|
|
Author(s):
|
Asuman Turkmen*+ and Nedret Billor
|
|
Companies:
|
Auburn University and Auburn University
|
|
Address:
|
Auburn University, 222 Parker Hall, Auburn, AL, 36849,
|
|
Keywords:
|
Partial least squares ; Outlier ; Robustness ; Multicollinearity
|
|
Abstract:
|
Partial Least Squares (PLS) regression is an alternative to ordinary least squares (OLS) regression for relating observed responses to a set of explanatory variables where the explanatory variables are highly collinear and where they outnumber the observations. Ordinary PLS regression is known to be very sensitive to outlying observations since it is based on maximizing the sample covariance matrix between the response and a set of explanatory variables. Therefore, in this study, a robust PLS method (RoPLS), which is resistant to masking and swamping problems, is proposed. We also explore the robustness properties of the proposed robust PLS method. Real and simulated data sets are used to compare the performance of the RoPLS with the existing methods.
|
