由離群值建構的基因分析

Abstract

這個計劃我們將討論在使用於基因研究之離群值平均（見 Tibshirani and Hastie(2007, Biostatistics) ）的無母數近似性質。其中我們將導出其近似分配，近 似平均及變異數。為了了解它在選取有致病基因的能力上，我們也將研究在某些分配下 的檢力模擬分析。實際資料之分析也會詳加討論。

This project plans to develop the asymptotic properties of the outlier mean (see Tibshirani and Hastie(2007, Biostatistics)) for gene expression analysis. As a statistic or an estimator, we will derive the asymptotic distribution of the outlier mean and study its asymptotic behavior in terms of asymptotic mean and variance. For evaluation of its performance for detection of influential genes, we will study the power for some tests constructed through asymptotic distribution of the outlier mean. In addition, data analyses will also be performed and the results will be displayed.