|
Activity Number:
|
374
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #303530 |
|
Title:
|
Bayesian Principal Component Regression with Data-Driven Components Selection
|
|
Author(s):
|
Liuxia Wang*+
|
|
Companies:
|
Sentrana Inc.
|
|
Address:
|
1725 Eye Street, Suite 900, Washington, DC, 20006,
|
|
Keywords:
|
Probabilistic principal component analysis ; Dynamic variable selection ; Dimensionality reduction
|
|
Abstract:
|
Principal component regression (PCR) has two steps: estimate the principal components and perform the regression using these components. These steps generally are performed sequentially. In PCR a crucial issue is the selection of the principal components to include in regression. In this paper, we build a hierarchical probabilistic PCR model with a dynamic component selection procedure. A latent variable is introduced to select promising subsets of components based upon the significance of the relationship between the response variable and principal components in the regression step. We illustrate this model using real and simulated examples. Our simulations demonstrate that our approach outperforms some existing methods in terms of rooted mean squared error (RMSE) of the regression coefficient.
|
