Statistics Colloquium
Location
Mathematics/Psychology : 103
Date & Time
March 28, 2014, 11:00 am – 12:00 pm
Description
SpeakerDr. DoHwan Park
Dr. DoHwan Park
Department of Mathematics and Statistics
University of Maryland, Baltimore County
Title
Empirical Null using Mixture Distributions and Its Application in Local False Discovery Rate
Abstract
When high dimensional data is given, it is often of interest to distinguish between significant (non-null, Ha) and non-significant (null, H0) group from mixture of two by controlling type I error rate. One popular way to control the level is the false discovery rate (FDR). This talk considers a method based on the local false discovery rate. In most of the previous studies, the null group is commonly assumed to be a normal distribution. However, if the null distribution can be departure from normal, there may exist too many or too few false discoveries (belongs null but rejected from the test) leading to the failure of controlling the given level of FDR. We propose a novel approach which enriches a class of null distribution based on mixture distributions. We provide real examples of gene expression data, fMRI data and protein domain data to illustrate the problems for overview.
Abstract
When high dimensional data is given, it is often of interest to distinguish between significant (non-null, Ha) and non-significant (null, H0) group from mixture of two by controlling type I error rate. One popular way to control the level is the false discovery rate (FDR). This talk considers a method based on the local false discovery rate. In most of the previous studies, the null group is commonly assumed to be a normal distribution. However, if the null distribution can be departure from normal, there may exist too many or too few false discoveries (belongs null but rejected from the test) leading to the failure of controlling the given level of FDR. We propose a novel approach which enriches a class of null distribution based on mixture distributions. We provide real examples of gene expression data, fMRI data and protein domain data to illustrate the problems for overview.
Tags: