方差未知时两样本正态混合模型齐-性的修正似然比检验

摘要

本文讨论了方差未知时检验两样本正态混合模型齐一性的修正似然比统计量的极限性质,证明了原假设下修正似然比统计量的渐近分布为自由度为1的卡方分布.%This paper investigates the asymptotic properties of a modified likelihood ratio statistics for testing homogeneity in normal mixture models of two samples with an unknown structural parameter. The asymptotic null distribution of the modified likelihood ratio test statistic is found to be X2,(1).