統計学輪講(第49回)

日時      2003年 2月10日(火)    15時〜16時40分 
場所      経済学部新棟3階第3教室
講演者    Prof. M.S. Srivastava (University of Toronto)
演題      Multivariate Theory For Analyzing High-Dimensional Data

概要
In this talk I give multivariate theory for analyzing high-dimensional
data.  Such data arise, for example, in DNA microarrays where there are
observations on thousands of genes but only on few subjects/patients.
Theory and methods for reducing the dimension and drawing inference from
them will be presented.  The inference problems include one-sample,
two-sample, and MANOVA tests.   A sample measure of distance between two
populations is defined.   This sample squared distance is used in classifying
an individual with p-vector observation into one of several multivariate
populations by minimum distance rule.


統計学輪講のスケジュールに戻る.


Tokyo University