Title: Principal Component Methodology: Recent Advances and the Challenge 
of High-Dimensional Data

	Trevor Park
	University of Florida

Abstract:

The past decade has seen renewed interest in principal component analysis, 
particularly when subject to penalization or restrictions and as applied 
to high-dimensional or functional data.  In this talk, I will briefly 
survey some recent methodology, highlighting features and limitations 
of various methods.  I will also discuss some new directions related to 
gradient descent algorithms (like LARS).