自助法在小样本数据均值估计中的应用

Research on Data Analysis of Small Sample Based on Bootstrap Method

摘要:

自助法(Bootstrap)是较好的处理小样本数据的方法,其无先验性及计算过程中只需要实际观测数据的优越性,使其广泛地应用于小样本数据处理。针对方法自身存在的缺陷,文章比较了目前自助法常用的两种抽样方式的优劣性,并采用改进的样本经验分布函数来解决这个问题,通过对均值的估计证实了方法的有效性。

Abstract:

Bootstrap method is a good method to deal with the data of small sample, which is only dependent on the observation and doesn’t need other assumption. For these reasons, it is adopted generally. But as to its self defects, two common ways to create new sample according to former sample were compared to get the better one. At the same time the improved empirical distruction function of sample was introduced to solve the problem, which was based on Bootstrap method. The emulation of example from this paper proved the validity of the improving method.