Functional Embedding in the Discriminant Analysis of Gene Expression Profiles

• Activity Number: 406
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Ping-Shi Wu*+
• Companies: Lehigh University
• Keywords: Functional Data Analysis ; Discriminant Ananlysis ; Gene Expression Profiling

Abstract

Functional embedding (FEM) technique exploits the interface between multivariate and functional data, aiming at borrowing strength across the sample through FDA techniques in order to resolve the difficulties caused by the high dimensionality p. The configuration defines the domain of a suitable function space, thus leading to the functional embedding of the high-dimensional data. FEM may then be followed by FDA approaches for downstream analysis. Specific application of FEM to gene expression profiling for discriminant analysis is exercised. Functional discriminant function based on functional multinomial logit regression is evaluated on several published gene expression array data sets and a mass spectrometric data, and is shown to compare favorably with various methods. This bases on joint works with Hans-Georg Müller, Swee The and Wei-Min Huang.