주성분방법을 이용한 유전자발현자료를 위한 향상된 분류기

초록

cDNA microarray which allows the monitoring of expression levels for thousands of genes simultaneously may lead to a more complete understanding of the molecular variation among tumors. A reliable and precise classification of tumors is a substance for successful cancer study. This paper compares the performance of several classifiers using gene expression profile of leukemia. It is shown that the principal component method proposed in this paper improves the performance of several classification methods. The performance is illustrated by error rates for test data sets.