论文部分内容阅读
针对成败型效应实验研究中二项分布未知参数(成功概率)的假设检验问题,在复杂假设条件下,提出了一种基于Bayesian验后概率的序贯检验方法,建立了检验的判别准则,给出了判别准则临界值的计算方法.在给定截尾实验次数的条件下,提出了一种截尾方案,建立了截尾判断方法.最后结合示例,对上述方法的应用过程进行了说明,并和现有方法进行了分析比较.结果表明:在一定先验信息的条件下,该方法给出的检验样本量远小于经典方法确定的检验样本量.
Aiming at the problem of hypothesis testing of unknown parameters (success probability) of two distributions in the experimental study of success or failure effects, a sequential test method based on post-Bayesian test probability is proposed under complex assumptions to establish the discriminant criterion of test A method of calculating the critical value of the discriminant criterion is given. Under the condition that the number of censored experiments is given, a censoring scheme is proposed and a truncated judgment method is established. Finally, the application process of the above method is illustrated with examples. The results show that under the condition of certain prior information, the test sample size given by this method is much smaller than the test sample size determined by the classical method.