摘要
讨论了q个多元正态总体中关于统计假设H0:μ1=μ2=…=μq=μ0,Σ1=Σ2=…=Σq=σ2I的检验问题,并以χ2 分布为基础的级数形式给出了似然比检验统计量的零分布及在与原假设相接近的某些备择假设下非零分布的一个渐近展开式.
In this paper, the problem for testing hypothesis H_0:μ_1=μ_2=…μ_q=μ_0, Σ_1=Σ_2=…=Σ_q=σ~2I is discussed in q multivariate normal populations. Asymptotic distributions of a certain function of likelihood ratio statistic are derived in a series of Chi-square distribution under certain alternatives which are close to the null hypothesis.
出处
《广州大学学报（自然科学版）》
CAS
2004年第3期193-199,共7页
Journal of Guangzhou University:Natural Science Edition
