Saved in:
Bibliographic Details
Main Authors: Fan, Yixuan, Jiao, Zhanyi, Wang, Ruodu
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2301.12480
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909324233670656
author Fan, Yixuan
Jiao, Zhanyi
Wang, Ruodu
author_facet Fan, Yixuan
Jiao, Zhanyi
Wang, Ruodu
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2301_12480
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Testing mean and variance by e-processes
Fan, Yixuan
Jiao, Zhanyi
Wang, Ruodu
Statistics Theory
Probability
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.
title Testing mean and variance by e-processes
topic Statistics Theory
Probability
url https://arxiv.org/abs/2301.12480