N-distance tests for a composite hypothesis of goodness of fit |
| |
Authors: | A. Bakshaev |
| |
Affiliation: | (1) Novosibirsk State Technical University, Novosibirsk, Russia; |
| |
Abstract: | In this paper, we consider an application of N-distance theory for testing composite hypotheses of goodness of fit. This work is a continuation of our research started in [A. Bakshaev, Goodness of fit and homogeneity tests on the basis of N-distances, J. Stat. Plan. Inference, 139(11):3750–3758, 2009; A. Bakshaev, Nonparametric tests based on N-distances, Lith. Math. J., 48(4):368–379, 2008]. Particular attention is paid to normality and exponentiality tests. A comparative Monte Carlo power study for the proposed criteria is provided. Different alternatives in uni- and bivariate cases are investigated. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|