arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2301.12480math.STmath.PRstat.TH
收藏

通过 e-processes 检验均值与方差

Testing mean and variance by e-processes

Yixuan Fan, Zhanyi Jiao, Ruodu Wang

AI总结:

本文研究非平稳数据的条件均值与方差检验问题,为四类具有指定均值、方差及分布形状条件的非参数复合假设构建 e-values 和 p-values,通过 e-processes 构建检验方法,并通过模拟和实证研究验证了其有效性。

摘要:

我们探讨了针对非平稳数据检验条件均值与条件方差的问题。我们为四类非参数复合假设构建了 e-values 和 p-values,这些假设具有指定的均值和方差以及对数据生成分布形状的其他条件。这些形状条件包括对称性、单峰性及其组合。利用所获得的 e-values 和 p-values,我们通过 e-processes(也称为 testing by betting)构建了检验,并构建了一些基于合并 p-values 的检验以作比较。尽管我们主要关注单边检验,但也研究了均值的双边检验。在几种设置下进行了模拟和实证研究,阐明了基于 e-processes 的方法的特征。

英文摘要:

We address the problem of testing conditional mean and conditional variance for non-stationary data. We build e-values and p-values for four types of non-parametric composite hypotheses with specified mean and variance as well as other conditions on the shape of the data-generating distribution. These shape conditions include symmetry, unimodality, and their combination. Using the obtained e-values and p-values, we construct tests via e-processes, also known as testing by betting, as well as some tests based on combining p-values for comparison. Although we mainly focus on one-sided tests, the two-sided test for the mean is also studied. Simulation and empirical studies are conducted under a few settings, and they illustrate features of the methods based on e-processes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑