控制犯两类错误后验概率的统计试验方案设计

THE DESIGN OF STATISTICAL TEST BY CONTROLLING POST-PROBABILITY OF TWO ERRORS

本文用ＢＡＹＥＳ理论提出了通过控制犯两类错误后验概率来设计试验方案的一种新方法，与以往的方法比较，该方法利用先验分布的信息减少了样本的容量。

Applying Bayes theory, a new method of designing statistical test by controlling postprobability of two errors is presented in this paper. Comparing with classical methods, information of prior-distribution is applied to reduce sample size.