TITLE: A General Framework for High-Dimensional Inference and Multiple Testing

ABSTRACT:

We consider the problem of how to control the measures of false scientific discoveries in high-dimensional models. Towards this goal, we focus on the uncertainty assessment for low dimensional components in high-dimensional models. Specifically, we propose a novel decorrelated likelihood based framework to obtain valid p-values for generic penalized M-estimators. Unlike most existing inferential methods which are tailored for individual models, our method provides a general framework for high-dimensional inference and is applicable to a wide variety of applications, including generalized linear models, graphical models, classifications and survival analysis. The proposed method provides optimal tests and confidence intervals. The extensions to general estimating equations are discussed. Finally, we show that the p-values can be combined to control the false discovery rate in multiple hypothesis testing.

Bio

Dr. Ning is a post-doctoral fellow in the department of Operations Research and Financial Engineering at the Princeton University. Prior to joining into the Princeton University, he was a visiting scholar in the department of Statistics and Actuarial Science at the University of Waterloo. He received his Ph.D in Biostatistics from the Johns Hopkins University. He has received numerous awards in statistics, including best student paper award from International Biometric Society and David Byar young investigator award from American Statistical Association. His research interests focus on the uncertainty assessment in data analysis with applications to biology, medicine and public health.