[Statlist] Research Seminar in Statistics *FRIDAY 28 SEPTEMBER 2018* GSEM, University of Geneva

gsem-support-instituts g@em-@upport-|n@t|tut@ @end|ng |rom un|ge@ch
Mon Sep 24 10:41:56 CEST 2018


Dear All,

We are pleased to invite you to our next Research Seminar.

Looking forward to seeing you

Organizers :                                                                                   
E. Cantoni - S. Engelke - D. La Vecchia - E. Ronchetti
S. Sperlich - F. Trojani - M.-P. Victoria-Feser

FRIDAY 28 SEPTEMBER 2018 at 11:15am, Uni-Mail M 5220

Testing for rotational symmetry against new hyperspherical distributions
Eduardo García PORTUGUÉS - Carlos III University of Madrid, Spain

ABSTRACT:

Motivated by the central role played by rotationally symmetric distributions in directional statistics, we consider the problem of testing rotational symmetry on the hypersphere. We adopt a semiparametric approach and tackle the situations where the location of the symmetry axis is either specified or unspecified. For each problem, we define two tests and study their asymptotic properties under very mild conditions.

The tests are aimed to detect location-like and scatter-like alternatives to rotational symmetry by relying on know uniformity tests for the hypersphere. We introduce two new classes of directional distributions that extend the rotationally symmetric class and are of independent interest and prove that each test is locally asymptotically maximin, in the Le Cam sense, for one kind of the alternatives given by the se new classes, both for specified and unspecified symmetry axis.

Conveniently, the tests can be combined into a hybrid test that is consistent against both alternatives, which we illustrate through the investigation of the presence of rotational symmetry in protein structures. If time allows, a new software package that collects several tests of uniformity for the hypersphere will be briefly discussed.


Visit the website: https://www.unige.ch/gsem/en/research/seminars/rcs/




More information about the Statlist mailing list