A Kolmogorov-Smirnov Test for r Samples

• Autorzy: Böhm W. , Hornik K.
• Języki publikacji: EN

Abstrakty

EN

We consider the problem of testing whether r ≥2 samples are drawn from the same continuous distribution F(x). The test statistic we will study in some detail is defined as the maximum of the circular differences of the empirical distribution functions, a generalization of the classical 2-sample Kolmogorov-Smirnov test to r ≥2 independent samples. For the case of equal sample sizes we derive the exact null distribution by counting lattice paths confined to stay in the scaled alcove Ar of the affineWeyl group Ar-1. This is done using a generalization of the classical reflection principle. By a standard diffusion scaling we derive also the asymptotic distribution of the test statistic in terms of a multivariate Dirichlet series. When the sample sizes are not equal the reflection principle no longer works, but we are able to establish a weak convergence result even in this case showing that by a proper rescaling a test statistic based on a linear transformation of the circular differences of the empirical distribution functions has the same asymptotic distribution as the test statistic in the case of equal sample sizes.

Słowa kluczowe

EN

Kolmogorov-Smirnov test; lattice path counting; reflection principle; affine Weyl groups; asymptotics distrubution

• Wydawca: IOS Press
• Czasopismo: Fundamenta Informaticae
• Rocznik: 2012
• Tom: Vol. 117, nr 1-4
• Strony: 103--125
• Opis fizyczny: Bibliogr. 21 poz., tab., wykr.

Twórcy

autor

Böhm W.

autor

Hornik K.

• Institute for Statistics and Mathematics, Vienna University of Economics and Business, Augasse 2-6, A 1091 Vienna, Austria,

Bibliografia

• [1] David, T. H.: A three-sample Kolmogorov-Smirnov test, The Annals of Mathematical Statistics, 29, 1958, 842-851.
• [2] Durbin, J.: Distribution Theory for Tests Based on the Sample Distribution Function, SIAM, Philadeliphia, 1973.
• [3] Filaseta,M.: A new method for solving a class of ballot problems, J. Combin. Theory, 39(A), 1985, 102-111.
• [4] Gessel, I. M., Zeilberger, D.: Random walk in a Weyl chamber, Proceedings of The American Mathematical Society, 115, 1992, 27-31.
• [5] Gnedenko, B. V., Korolyuk, V. S.: On themaximumdiscrepancy between two empirical distributions (English translation), Doklady Akademii Nauk SSSR, 80, 1951, 525-528.
• [6] Grabiner, D. J.: Random walk in an alcove of an affine Weyl group, and non-colliding random walks on an interval, Journal of Combinatorial Theory (A), 97, 2002, 285-306.
• [7] Hájek, J., ˇSidák, Z.: Theory of Rank Tests, Academic Press, 1967.
• [8] Handa, B. R., Mohanty, S. G.: Higher dimensional lattice paths with diagonal steps, Disc. Math., 15, 1976, 137-140.
• [9] Karlin, S. P., McGregor, G.: Coincidence Probabilities, Pacific J. Math, 9, 1959, 1141-1164.
• [10] Kiefer, J.: Distance tests with good power for the nonparametric k-sample problem, (abstract), The Annals of Mathematical Statistics, 26, 1955, 775.
• [11] Kiefer, J.: K-sample analogues of the Kolmogorov-Smirnov and Cramér-v. Mises tests, The Annals of Mathematical Statistics, 30, 1959, 420-447.
• [12] Krattenthaler, C.: q-generalization of a ballot problem, Discrete Math, 126, 1994, 195-208.
• [13] Krattenthaler, C.: Asymptotics for random walks in alcoves of affine Weyl groups, Séminaire Lotharingien de Combinatoire, 52, 2007, Article B25i.
• [14] Kreweras, G.: Sur une classe de problemes de denombrement liés au treilles des partitions des entiers, Cahiers du Bur. Univ. de Rech. Oper., 6, 1965, 5-105.
• [15] Mohanty, S. G.: Lattice Path Counting and Applications, Academic Press, New York, 1979.
• [16] Ozols, V.: Generalization of the theorem of Gnedenko-Korolyuk to three samples in the case of two-sided boundaries, Latvijas PSR Zinatnu Akademijas Vestis, 10, 1956, 141-152.
• [17] Shorack, G. R., Wellner, J. A.: Empirical processes with applications to statistics, New York,Wiley, 1986.
• [18] Steck, G. P.: The Smirnov two-sample tests as rank tests, Ann. Math. Stat., 40, 1969, 1449-1466.
• [19] Takács, L.: On a three-sample test, Lecture Notes in Statistics (C. C. Heyde, Y. V. Prohorov, R. Pyke, S. T. Rachev, Eds.), 114, 1996.
• [20] Watanabe, T., Mohanty, S. G.: On an inclusion-exclusion formula based on the reflection principle, Discrete Math, 64, 1987, 281-288.
• [21] Zeilberger, D.: André's reflection proof generalized to the many-candidate ballot problem., Discrete Math, 44, 1983, 325-326.