基于序贯估计方法的电子器件技术参数统计推断

The statistical inference of the electronic device's technical parameters based on sequential estimation method

对A.瓦尔德所研究的序贯抽样方案,用序贯估计方法,研究A公司一类新投产的电子器件的特征参数并给出统计推断.利用这种方法对某些重要参数,只要恰当地选取样本容度,除了节省抽样量之外,还可提高推断的可靠程度及精确程度.进一步用两个定理证明了样本最佳密度、特征参数及置信区间等几个重要问题.

With sequential sampling program studied by A. Wald, sequential estimation methods are used to research the characteristic parameters of a new kind of electronic devices of company A （see main text） and statistical inference is given. The use of this method for some important parameters, as long as the proper degree of sampling capacity, in addition to saving the volume of samplings, can also improve the reliability and precision of the inference. Further two theorems are used to prove several important issues, such as the best density of the samples, the characteristic parameters and the confidence interval and so on.