|
Activity Number:
|
139
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 4, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #301879 |
|
Title:
|
Outlier Detection Using the Smallest Kernel Principal Components with Radial Basis Function Kernel
|
|
Author(s):
|
Yan Shen+ and Alan J. Izenman*+
|
|
Companies:
|
Temple University and Temple University
|
|
Address:
|
1136 Easton Ave, Somerset, NJ, 08873, Department of Statistics, Philadelphia, PA, 19122-6083,
|
|
Keywords:
|
Kernel Method ; Principal Component Analysis ; Outlier Detection
|
|
Abstract:
|
The smallest principal components have not attracted much attention in the statistics literature. This apparent lack of interest is due to the fact that, compared with the largest principal components that contain most of the total variance in the data, the smallest principal components only contain the noise of the data and, therefore, appear to contribute minimal information. This article proposes a method for outlier detection using the smallest kernel principal components in a feature space induced by radial basis function kernel. We show that the eigenvectors corresponding to the smallest kernel principal components can be viewed as that for which the residual sum of squares is minimized, and we can use those components to identify outliers with simple graphical techniques. Simulation studies show that under univariate outlier situation, the proposed method is as good as others.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2008 program |
