摘要
对于高维MANOVA假设检验问题,文章提出了一种新的检验统计量,且对数据的维数和个数的关系没有限制;在一般条件下,分别得到了新检验统计量在原假设和局部备择假设下的渐近分布.模拟结果表明新的检验具有一定的优势.
To test the equality of the mean vectors in MANOVA with high dimensional data, atest was proposed. The test permits arbitrary relationship between the data size and the datadimension. The asymptotic distributions were obtained in the null and non-null hypotheses undermild conditions. Simulation results illustrated that the proposed test is superior to the existingtest in some cases.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2015年第8期868-871,共4页
Transactions of Beijing Institute of Technology
基金
国家自然科学基金资助项目(11471035,11471030)
安徽省自然科学基金资助项目(1508085QA11)
关键词
高维数据
MANOVA
假设检验
渐近分布
high-dimension data
MANOVA
testing the hypotheses
asymptotic distributions