Optimal rank-based tests for homogeneity of scatter

Details Optimal rank-based tests for homogeneity of scatter Optimal rank-based tests for homogeneity of scatter

2008-06-18

arxiv.org/abs/0806.2963v1

Annals of Statistics 2008, Vol. 36, No. 3, 1261-1298

Mathematics

Statistics Theory

Published in at http://dx.doi.org/10.1214/07-AOS508 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of

Scientific paper

10.1214/07-AOS508

We propose a class of locally and asymptotically optimal tests, based on multivariate ranks and signs for the homogeneity of scatter matrices in $m$ elliptical populations. Contrary to the existing parametric procedures, these tests remain valid without any moment assumptions, and thus are perfectly robust against heavy-tailed distributions (validity robustness). Nevertheless, they reach semiparametric efficiency bounds at correctly specified elliptical densities and maintain high powers under all (efficiency robustness). In particular, their normal-score version outperforms traditional Gaussian likelihood ratio tests and their pseudo-Gaussian robustifications under a very broad range of non-Gaussian densities including, for instance, all multivariate Student and power-exponential distributions.

Affiliated with

Also associated with

No associations

Rating

Optimal rank-based tests for homogeneity of scatter does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Optimal rank-based tests for homogeneity of scatter, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Optimal rank-based tests for homogeneity of scatter will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-272721