一类混合模型的似然比检验
Likelihood Ratio Test of a Class of Mixture Model
-
摘要: 考虑一类混合模型g(x)=(1-ζ1-ζ2)f(x,θ0')+ζ1f(x,θi')+ζ2f(x,θ2'),其中ζ1,ζ2属于区域Ω:0≤ζ1,ζ2,ζ1+ζ2≤1,f为一给定的概率密度函数,θi'(i=0,1,2)为m维常数向量,本文讨论了该模型的似然比统计量的极限分布.Abstract: In this paper, a class of mixture model g(x)=(1-ζ1-ζ2)f(x,θ0')+ζ1f(x,θi')+ζ2f(x,θ2'),is discussed, where 0≤ζ1,ζ2,ζ1+ζ2≤1, f is a probability density function, and θi'(i=0,1,2) is a m-dimensional constant vector. The limit distribution of likelihood ratio statistic of the model is given.