Online Program

WITHDRAWN: Sparse Principal Component Analysis in Linear Regression Survival Model

Guanyu Hu, Florida State University 

Keywords: Survival analysis, Microarray data, Sparse Principal Component Analysis, Linear regression model

Cox proportional hazards model plays an important role in the covariates of the traditional survival analysis. But in contemporary biostatistics study, Cox model doesn't do well in many research objects like microarray data. In this paper, we study the linear regression survival model, which analyzes the relationship between the covariates and event time transformed monotonically, like logarithm. Compared with the traditional regression model, the regression model we study is with the high-dimensional survival data, which is right censored. And we use the sparse principal component analysis (SPCA), which is a very famous method in high-dimensional data analysis to deal with this kind of data. SPCA can produce a principal component that involves the genes of limit numbers. SPCA regression model can give us an alternative to Cox model in survival analysis for microarray data.