Multiple Decision Functions and High-Dimensional Failure Time Data

• Activity Number: 463
• Type: Topic Contributed
• Date/Time: Wednesday, August 4, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Edsel A. Pena*+ and Joshua D. Habiger and Wensong Wu
• Companies: University of South Carolina and National Agricultural Statistics Service/National Institute of Statistical Sciences and University of South Carolina
• Keywords: Multiple testing; high-dimensional data; family wise error rate; false discovery rate.

Abstract

In this talk we present exploratory ideas and new procedures for multiple decision problems using high-dimensional failure data arising in biology, medicine, economics, and engineering. Of interest is how to choose the important explanatory variables from a collection of variables to predict an event time. The decision will be based on, possibly right-censored, event times. In such problems it is imperative that the issue of multiplicity be considered. Consequently, the multiple decision functions of interest are those which control, either weakly or strongly, family wise error rate (FWER) and false discovery rate (FDR). We describe general classes of multiple decision functions controlling FWER and FDR. We also partially address optimality issues.