P. Navarro Esteban, E. García Portugués, J. A. Cuesta Albertos

We study a projection-based class of uniformity tests on the hypersphere using statistics that integrate, along all possible directions, the weighted quadratic discrepancy between the empirical cumulative distribution function of the projected data and the projected uniform distribution. Simple expressions for several test statistics are obtained in particular dimensions. Despite their different origins, the proposed class and the well-studied Sobolev class of uniformity tests are shown to be related. Our new parametrization proves itself advantageous by allowing to derive new tests for hyperspherical data that neatly extend the circular tests by Watson, and Rothman, and by introducing the first instance of an Anderson-Darling-like test in such context. The asymptotic distributions and the local optimality against certain alternatives of the new tests are obtained. A simulation study corroborates the theoretical findings and a real data example illustrates the usage of the new tests.

Keywords: Circular data, directional data, hypersphere, Sobolev tests, uniformity

Scheduled

ENP-1 Non Parametric Statistics
September 4, 2019  2:45 PM
I3L9. Georgina Blanes building


Other papers in the same session

Evaluación del efecto de las covariables en las curvas ROC

A. Fanjul Hevia, W. González Manteiga, J. C. Pardo Fernández

Generalización de la curva ROC sobre combinaciones lineales

S. Pérez Fernández, P. Martínez Camblor, N. Corral Blanco


Cookie policy

We use cookies in order to be able to identify and authenticate you on the website. They are necessary for the correct functioning of it, and therefore they can not be disabled. If you continue browsing the website, you are agreeing with their acceptance, as well as our Privacy Policy.

Additionally, we use Google Analytics in order to analyze the website traffic. They also use cookies and you can accept or refuse them with the buttons below.

You can read more details about our Cookie Policy and our Privacy Policy.