|
Activity Number:
|
506
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #309856 |
|
Title:
|
An Improved Classification Procedure for High-Dimension and Highly Noisy Data Using Data Defined Distance and Comparison Test
|
|
Author(s):
|
Shuang Liu*+ and David Rocke
|
|
Companies:
|
University of California, Davis and University of California, Davis
|
|
Address:
|
404 Russell Park Apt 3, Davis, CA, 95616,
|
|
Keywords:
|
|
|
Abstract:
|
Due to the high dimensionality and existence of significant noises in microarray datasets, the differences between gene groups are not well explained by many classification approaches. In this study of differentially regulated gene groups and pathways, five classification methods were first applied to each two-group classification individually. Misclassification rates were then estimated using cross-validation. Finally, a permutation test was used to improve the evaluation of the classification result by comparing the misclassification rates. By this means, we were able to reveal the latent differences between gene groups. The permutation test also showed that the Mahalanobis distance classifier out-performed the other classification methods, and the K-nearest neighbor method based on the correlation distance appeared to be a better solution to the problem of high-dimensionality.
|
- 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 2007 program |
