[Ml-stat-talks] Statistics Seminar this Wednesday

Jelena Bradic jbradic at Princeton.EDU
Tue Dec 7 10:06:16 EST 2010



Dear all,

This coming Wednesday  at 1:30pm in Statistics Lab (Sherrerd 213) we  
will be having a Seminar meeting.



Speaker:
Jelena Bradic

Title:
Regularization for Cox's Proportional Hazards Model with NP  
dimensionality


Abstract:

High throughput genetic sequencing arrays with thousands of  
measurements per
sample and a great amount of related censored clinical data have  
increased demand-
ing need for better measurement speci c model selection. In this  
paper we estab-
lish strong oracle properties of non-concave penalized methods for  
non-polynomial
(NP) dimensional data with censoring in the framework of Cox's  
proportional haz-
ards model. A class of folded-concave penalties are employed and both  
LASSO and
SCAD are discussed speci cally. We unveil the question under which  
dimensionality
and correlation restrictions can an oracle estimator be constructed  
and grasped. It
is demonstrated that non-concave penalties lead to signi cant  
reduction of the "ir-
representable condition" needed for LASSO model selection  
consistency. The large
deviation result for martingales, bearing interests of its own, is  
developed for charac-
terizing the strong oracle property. Moreover, the non-concave  
regularized estimator,
is shown to achieve asymptotically the information bound of the  
oracle estimator.
A coordinate-wise algorithm is developed for nding the grid of  
solution paths for
penalized hazard regression problems, and its performance is  
evaluated on simulated
and gene association study examples


Best,
Jelena


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cs.princeton.edu/pipermail/ml-stat-talks/attachments/20101207/3ec54b69/attachment.html>


More information about the Ml-stat-talks mailing list