项目名称: 超高维数据中若干检验问题的研究
项目编号: No.11501092
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 数理科学和化学
项目作者: 冯龙
作者单位: 东北师范大学
项目金额: 18万元
中文摘要: 高维数据分析是近年来统计学研究中的热点。随着科技的发展,实际中收集的一些数据的维数越来越高,大大的超过了样本量,比如基因数据,航空数据,金融数据等。由于维数比样本量大,这就导致传统的统计分析方法失效。这就需要新的统计分析法来处理高维数据的相关问题。本项目将深入研究高维数据中若干检验问题,包括高维单样本均值检验,高维两样本均值检验,高维球形检验等。现有的很多方法都是基于发散因子模型或者多元正态分布假设下开发的,从而并不稳健。本项目基于空间符号统计量、空间秩统计量等方法期望可以提出一些新的检验方法,并应用于实际领域。
中文关键词: 假设检验;空间符号统计量;空间秩统计量
英文摘要: With the rapid development of technology, various types of high-dimensional data have been generated in many areas, such as hyperspectral imagery, internet portals, microarray analysis and DNA. Many traditional methods may not work anymore in this situation since they assume that p keeps unchanged as n increases. This challenge calls for new statistical tests to deal with high-dimensional data. In this project, we consider some study on high dimensional test problems, such as one sample location test, two sample location tests, sphericity test. Nowadays, many methods are all based on the assumption of diverging factor model or multivariate normal assumption, which is not robust and efficient. We aim to propose some new test procedures based on spatial sign and spatial rank and use them in practice.
英文关键词: hypothesis test;spatial sign;spatial rank