On two-stage Monte Carlo tests of composite hypotheses

被引：11

作者：

Baddeley, Adrian ^{[1
,3
]}

Hardegen, Andrew ^{[2
]}

Lawrence, Thomas ^{[2
]}

Milne, Robin K. ^{[2
]}

Nair, Gopalan ^{[2
,3
]}

Rakshit, Suman ^{[1
]}

机构：

[1] Curtin Univ, Dept Math & Stat, GPO Box U1987, Perth, WA 6845, Australia

[2] Univ Western Australia, Sch Math & Stat, 35 Stirling Hwy, Crawley, WA 6009, Australia

[3] CSIRO, Data61, Leeuwin Ctr, 65 Brockway Rd, Floreat, WA 6014, Australia

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 2017年 / 114卷

基金：

澳大利亚研究理事会;

关键词：

Conservative test; Dao-Genton test; Double bootstrap; Nested bootstrap; p-value; NUISANCE PARAMETER; P-VALUES; INFERENCE;

D O I：

10.1016/j.csda.2017.04.003

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

A major weakness of the classical Monte Carlo test is that it is biased when the null hypothesis is composite. This problem persists even when the number of simulations tends to infinity. A standard remedy is to perform a double bootstrap test involving two stages of Monte Carlo simulation: under suitable conditions, this test is asymptotically exact for any fixed significance level. However, the two-stage test is shown to perform poorly in some common applications: for a given number of simulations, the test with the smallest achievable significance level can be strongly biased. A 'balanced' version of the two-stage test is proposed, which is exact, for all achievable significance levels, when the null hypothesis is simple, and which performs well for composite null hypotheses. (C) 2017 Elsevier B.V. All rights reserved.

引用

页码：75 / 87

页数：13

共 50 条

[1] A Monte Carlo procedure for two-stage tests with correlated data
Martin, ER
Kaplan, NL
GENETIC EPIDEMIOLOGY, 2000, 18 (01) : 48 - 62
[2] Monte Carlo based techniques of two-stage phase shifting algorithms
Miranda, M.
Dorrio, B. V.
OPTICS AND LASERS IN ENGINEERING, 2011, 49 (03) : 439 - 444
[3] Two-Stage Monte Carlo Tree Search for Connect6
Yen, Shi-Jim
Yang, Jung-Kuei
IEEE TRANSACTIONS ON COMPUTATIONAL INTELLIGENCE AND AI IN GAMES, 2011, 3 (02) : 100 - 118
[4] COMBINING MONTE-CARLO AND COX TESTS OF NONNESTED HYPOTHESES
SCHORK, N
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 1993, 22 (04) : 939 - 954
[5] Probabilistic estimates of scenario loss through a two-stage Monte Carlo approach
Ioannou, Ioanna
Chryssanthopoulos, Marios K.
APPLICATIONS OF STATISTICS AND PROBABILITY IN CIVIL ENGINEERING, 2011, : 252 - 259
[6] Are Quasi-Monte Carlo algorithms efficient for two-stage stochastic programs?
H. Heitsch
H. Leövey
W. Römisch
Computational Optimization and Applications, 2016, 65 : 567 - 603
[7] A survey of Monte Carlo algorithms for maximizing the likelihood of a two-stage hierarchical model
Booth, James G.
Hobert, James P.
Jank, Wolfgang
STATISTICAL MODELLING, 2001, 1 (04) : 333 - 349
[8] Are Quasi-Monte Carlo algorithms efficient for two-stage stochastic programs?
Heitsch, H.
Leoevey, H.
Roemisch, W.
COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2016, 65 (03) : 567 - 603
[9] An investigation of two-stage tests
Vandemeulebroecke, Marc
STATISTICA SINICA, 2006, 16 (03) : 933 - 951
[10] The geometry of two-stage tests
Proschan, MA
STATISTICA SINICA, 2003, 13 (01) : 163 - 177

← 1 2 3 4 5 →