On the automatic selection of the tuning parameter appearing in certain families of goodness-of-fit tests. (English) Zbl 07193810

Tenreiro, Carlos

J. Stat. Comput. Simulation 89, No. 10, 1780-1797 (2019).

Summary: The situation, common in the current literature, is that of a whole family of location-scale/scale invariant test statistics, indexed by a parameter \(\lambda \in \Lambda\), is available to test the goodness of fit of \(F\), the underlying distribution function of a set of independent real-valued random variables, to a location-scale/scale family of distribution functions. The power properties of the tests associated with the different statistics usually depend on the parameter \(\lambda\), called the “tuning parameter”, which is the reason that its choice is crucial to obtain a performing test procedure. In this paper, we address the automatic selection of the tuning parameter when \(\Lambda\) is finite, as well as the calibration of the associated goodness-of-fit test procedure. Examples of existing and new tuning parameter selectors are discussed, and the methodology presented of combining different test statistics into a single test procedure is applied to well known families of test statistics for normality and exponentiality. A simulation study is carried out to access the power of the different tests under consideration, and to compare them with the fixed tuning parameter procedure, usually recommended in the literature.

MSC:

62G10	Nonparametric hypothesis testing
62G20	Asymptotic properties of nonparametric inference

Keywords:

goodness-of-fit tests; data-based tuning parameter selection; calibration; normality tests; empirical characteristic function; exponentiality tests; empirical Laplace transform

Software:

R

References:

[1]	Epps TW, Pulley LB.A test for normality based on the empirical characteristic function. Biometrika. 1983;70:723-726. doi: 10.1093/biomet/70.3.723[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0523.62045
[2]	Baringhaus L, Henze N.A class of consistent tests for exponentiality based on the empirical Laplace transform. Ann Inst Statist Math. 1991;43:551-564. doi: 10.1007/BF00053372[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0760.62040
[3]	Henze N.A new flexible class of omnibus tests for exponentiality. Commun Statist Theory Methods. 1993;22:115-133. doi: 10.1080/03610929308831009[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 0777.62047
[4]	Güntler N, Henze N.Goodness-of-fit tests for the Cauchy distribution based on the empirical characteristic function. Ann Inst Statist Math. 2000;52:267-286. doi: 10.1023/A:1004113805623[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0959.62041
[5]	Klar B.Goodness-of-fit tests for the exponential and the normal distribution based on the integrated distribution function. Ann Inst Statist Math. 2001;53:338-353. doi: 10.1023/A:1012422823063[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1027.62026
[6]	Henze N, Meintanis SG.Test of fit for exponentiality based on the empirical Laplace transform. Statistics. 2002;36:147-161. doi: 10.1080/02331880212042[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1008.62044
[7]	Meintanis SG.Goodness-of-fit tests for the logistic distribution based on empirical transforms. Sankhya Ser A. 2004;66:306-326. [Google Scholar] · Zbl 1193.62080
[8]	Meintanis SG.A class of omnibus tests for the Laplace distribution based on the empirical characteristic function. Commun Statist Theory Methods. 2004;33:925-948. doi: 10.1081/STA-120028735[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1066.62049
[9]	Meintanis S, Swanepoel J, Allison J.The probability weighted characteristic function and goodness-of-fit testing. J Statist Plann Inference. 2014;146:122-132. doi: 10.1016/j.jspi.2013.09.011[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1279.62107
[10]	Fan Y.Goodness-of-fit tests based on kernel density estimators with fixed smoothing parameters. Econ Theory. 1998;14:604-621. doi: 10.1017/S0266466698145036[Crossref], [Google Scholar]
[11]	Henze N, Zirkler B.A class of invariant consistent tests for multivariate normality. Commun Statist Theory Methods. 1990;19:3595-3617. doi: 10.1080/03610929008830400[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 0738.62068
[12]	Tenreiro C.On the role played by the fixed bandwidth in the Bickel-Rosenblatt goodness-of-fit test. SORT. 2005;29:201-216. [Google Scholar] · Zbl 1274.62333
[13]	Tenreiro C.On the finite sample behaviour of fixed bandwidth Bickel-Rosenblatt test for univariate and multivariate uniformity. Commun Statist Simulation Comput. 2007;36:827-846. doi: 10.1080/03610910701418697[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1126.62036
[14]	Tenreiro C.On the choice of the smoothing parameter for the BHEP goodness-of-fit test. Comput Statist Data Anal. 2009;53:1038-1053. doi: 10.1016/j.csda.2008.09.002[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1452.62322
[15]	Kawata T. Fourier analysis in probability theory. New York: Academic Press; 1972. [Google Scholar] · Zbl 0271.60022
[16]	Feller W. An introduction to probability theory and its applications. New York: John Wiley; 1971. [Google Scholar] · Zbl 0219.60003
[17]	Fromont M, Laurent B.Adaptive goodness-of-fit tests in a density model. Ann Statist. 2006;34:680-720. doi: 10.1214/009053606000000119[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1096.62040
[18]	Tenreiro C.An affine invariant multiple test procedure for assessing multivariate normality. Comput Statist Data Anal. 2011;55:1980-1992. doi: 10.1016/j.csda.2010.12.004[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1328.62357
[19]	Tenreiro C.A new test for multivariate normality by combining extreme and non-extreme BHEP test statistics. Commun Statist Simulation Comput. 2017;46:1746-1759. doi: 10.1080/03610918.2015.1011334[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1364.62109
[20]	Dudoit S, van der Lann MJ. Multiple testing procedures with applications to genomics. New York: Springer; 2008. [Crossref], [Google Scholar] · Zbl 1261.62014
[21]	Allison J, Santana L.On a data-dependent choice of the tuning parameter appearing in certain goodness-of-fit tests. J Stat Comput Simul. 2015;85:3276-3288. doi: 10.1080/00949655.2014.968781[Taylor & Francis Online], [Web of Science ®], [Google Scholar]
[22]	R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2014. Available from: http://www.R-project.org/. [Google Scholar]
[23]	Baringhaus L, Henze N.A consistent test for multivariate normality based on the empirical characteristic function. Metrika. 1988;35:339-348. doi: 10.1007/BF02613322[Crossref], [Google Scholar] · Zbl 0654.62046
[24]	Baringhaus L, Danschke R, Henze N.Recent and classical tests for normality. A comparative study. Commun Statist Simulation Comput. 1989;18:363-379. doi: 10.1080/03610918908812764[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 0695.62120
[25]	Arcones MA, Wang Y.Some new tests for normality based on U-processes. Statis Probab Lett. 2006;76:69-82. doi: 10.1016/j.spl.2005.07.003[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1085.62047
[26]	Romão X, Delgado R, Costa A.An empirical power comparison of univariate goodness-of-fit tests for normality. J Stat Comput Simul. 2010;80:545-591. doi: 10.1080/00949650902740824[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1195.62056
[27]	Yap BW, Sim CH.Comparisons of various types of normality tests. J Stat Comput Simul. 2011;81:2141-2155. doi: 10.1080/00949655.2010.520163[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1431.62201
[28]	Henze N, Meintanis SG.Recent and classical tests for exponentiality: a partial review with comparisons. Metrika. 2005;61:29-45. doi: 10.1007/s001840400322[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1062.62076
[29]	Janssen A.Global power functions of goodness of fit tests. Ann Statist. 2000;28:239-253. doi: 10.1214/aos/1016120371[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1106.62329
[30]	Lehmann EL, Romano JP. Testing statistical hypotheses. New York: Springer; 2005. [Google Scholar] · Zbl 1076.62018
[31]	Shorack GR, Wellner JA. Empirical processes with applications to statistics. New York: John Wiley; 1986. [Google Scholar] · Zbl 1170.62365
[32]	Carathéodory C. Theory of functions of a complex variable. New York: Chelsea; 1983. [Google Scholar]