Let x1,…, xn be the life spans of n items. Suppose that we start the experi-ment at the time t=0 for all the items simultaneuously aud stop the experiment at the timet=r, where r is a stopping time. The observed data set is Zn=(x1,…, xk, r(x)),where xi(i=1,…, n) is the order statistic of xi, i=1,…, n and k=k(x) is the num-ber of the observed data. Suppose that the distribution family of xi, i=1,…, n is i.i.d.exponential with life expectation θ0. For testing the hypothesis H0: θ≤θ0 againstH1: θθ0, we use the total time of experiment Sr=sub from i=1 to k xi +(n-k)r (x) as the teststatistic. We reject H0 as large value of Sr is observed. In this paper for a given stoppingtime r we construct a stopping time ro so that the resulting test is a best improvement ofthe one, which is based on r, in the sense that the power functions of the two tests are thesame but the test based on r0 is the one which has the minimnm total experiment time.
展开▼
机译:让x 1 sub>,...,x n sub>是n项的寿命。假设我们在时间t = 0开始实验,对于同时的所有项目,在TIMET = R处停止实验,其中R是停止时间。观察到的数据集是z n sub> =(x 1 sub>,...,x k sub>,r(x)),其中x i sub>(i = 1,...,n)是x i sub>的顺序统计,i = 1,...,n和k = k(x)是观察到的num-ber数据。假设分布族的x i sub>,i = 1,...,n是i.i.d.exponential with life期望θ> 0。用于测试假设H 0 sub>:θ≤θ 0 sub>反对H 1 sub>:θ>θ 0 sub>,我们使用实验的总时间S R sub> = sub从i = 1到kx i sub> +(nk)r(x)作为Teststatic。我们拒绝H 0 sub>,因为观察到S R sub>的大值。在本文中,对于给定的停止时间R,我们构建一个停止时间RO,使得所得到的测试是基于R的最佳改进,这是两个测试的功率函数而是基于测试的功率功能而是基于R 0 sub>是具有全部实验时间的最小值的。
展开▼