B1037
Title: On directional runs tests and their local and asymptotic optimality properties
Authors: Maxime Boucher - Universite Libre de Bruxelles (ULB) (Belgium) [presenting]
Thomas Verdebout - Universite Libre de Bruxelles (Belgium)
Yuichi Goto - Kyushu University (Japan)
Abstract: The problem of testing the randomness of directional data is tackled. First, a new concept is defined of runs properly adapted to the directional context. Then, it is shown that tests based on the later runs enjoy some local and asymptotic properties against local alternatives with serial dependence. The finite sample performances of the tests are computed using Monte Carlo simulations and their usefulness is shown on a real data illustration that involves the analysis of sunspots on the Photosphere.