Paper Infomation
Cancer Gene Extraction Based on Ideal Gene Model
Full Text(PDF, 624KB)
Author: Jie Ni, Fan Wu, Meixiang Jin, Yixing Bai, Yunfei Guo
Abstract: In this paper, we exclude 1561 unrelated genes through the definition of weighted distance, and then we remove 252 genes by Pearson's correlation coefficient. Finally, we gain the best combination of genes which includes five genes through cluster analysis and discriminant analysis; dimension reduction from 2000 dimensions to 5 dimensions is completed.
Keywords: Ideal Gene; Cluster Analysis; Feature Gene
References:
[1] WANG SHOU-JUE，ZHOU LING-FEl. Gene Selection for Gene Expression Data Analysis, CONTROL & AUTOMATION, Vol 24,2008.
[2] LI Ying-xin, RUAN Xiao-gang. Cancer Subtype Recognition and Feature Selection with Gene Expression Profiles, ACTA ELECTRONICA SINICA,Vol 33,2005.