Orbit-based conditional tests. A link between permutations and Markov bases

被引：0

作者：

Fontana, Roberto ^{[1
]}

Crucinio, Francesca Romana ^{[2
]}

机构：

[1] Politecn Torino, Dept Math Sci, Turin, Italy

[2] Univ Warwick, Dept Stat, Coventry, W Midlands, England

来源：

JOURNAL OF STATISTICAL PLANNING AND INFERENCE | 2020年 / 205卷

关键词：

Algebraic statistics; Conditional test; Discrete exponential family; Markov chain Monte Carlo; Permutation test;

D O I：

10.1016/j.jspi.2019.05.007

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

Algebraic sampling methods are a powerful tool to perform hypothesis testing for non-negative discrete exponential families, when the exact computation of the test statistic null distribution is computationally infeasible. We propose an improvement of the accelerated sampling described by Diaconis and Sturmfels (1998) based on permutations. We thus establish a link between standard permutation and algebraic-statistics-based sampling. We prove that the permutations-based sampling gives the lowest approximation errors and we validate our algorithm through a simulation study on three applications (data fitting, two sample tests and linear regression). (C) 2019 Elsevier B.V. All rights reserved.

引用

页码：23 / 33

页数：11