摘要
设(?)～N_p((?)_j,(?)_j),S_j～W_p(Σ_j,n_j),(j=1,2,…,k)且相互独立。设原假设为H_1:Σ+1=Σ_2=…=Σ_k且(?)_1=(?)_2=…=(?)_k=(?) 设(?)_j～CN_p((?)_j,Q_j),A_j～CW_p(Q_j,n_j)(j=1,2,…,k)且相互独立。设原假设为H_2:Q_1=Q_2=…=Q_k且(?)_1=(?)_1=…=(?)_k=(?) 本文讨论了以上两个检验问题,给出了其似然比统计量在原假设为真时的累积分布函数的渐近展开式。
Let (?)～N_p((?), ∑_j), S_j～W_p(∑_j, n_j) (j=1, 2, …, k) and let them be mutually independent. Let the null hypothesis beH_1: ∑_1=∑_2=……=∑_k and (?)=(?)=……=(?)=(?)Let (?)～CN_p, ((?), Q_1), A_j～CW_p(Q_j, n_j)(j=1, 2, …, k) and let them be independent. Let the null hypothesis beH_2: Q_1=Q_2=……Q_k and (?)=(?)=……=(?)=(?)In this article the asympotic expansions of their likelihood ratio statistics under the null hypotheses are given.
出处
《湘潭大学自然科学学报》
CAS
CSCD
1990年第3期14-21,共8页
Natural Science Journal of Xiangtan University
