Regularized Supervised Principal Component Analysis

• Activity Number: 332
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: IMS
• Author(s): Gen Li*+ and Haipeng Shen and Jianhua Z. Huang
• Companies: and University of North Carolina at Chapel Hill and Texas A&M
• Keywords: Supervised dimension reduction ; Regularized principal component analysis ; Reduced rank regression ; SupSVD ; SupSFPC

Abstract

We introduce a supervised principal component analysis model for supervised dimension reduction where the low rank structure of the data of interest is potentially driven by additional variables measured on the same set of samples. It can make use of the information in the additional variables to accurately extract underlying structures that are more interpretable. The model is formulated in a hierarchical fashion using latent variables, and a modified expectation-maximization algorithm for parameter estimation is developed, which is computationally efficient. The asymptotic properties for the estimated parameters are derived. We further adapt the model to functional and sparse situations, where the special features are imposed through regularization. We use comprehensive simulations and real data examples to illustrate the advantages of the proposed method.