DISCUSSION

Hallin, Paindaveine and Siman -- hereafter HPS -- are to be congratulated for the appreciation their paper is receiving from The Annals of Statistics. The authors personally are indebted for the attention they devoted to their paper, -- hereafter KM. The topic is quite delicate and multifaceted, hen...

Full description

Saved in:
Bibliographic Details
Published in:The Annals of statistics 2010-04, Vol.38 (2), p.685
Main Authors: Kong, Linglong, Mizera, Ivan
Format: Article
Language:English
Subjects:
Multivariate analysis
Regression analysis
Studies
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Hallin, Paindaveine and Siman -- hereafter HPS -- are to be congratulated for the appreciation their paper is receiving from The Annals of Statistics. The authors personally are indebted for the attention they devoted to their paper, -- hereafter KM. The topic is quite delicate and multifaceted, hence all the misunderstandings they allege show deficiencies of their exposition rather than anything else. This article discusses certain aspects of using quantiles to obtain insights about multivariate data, suggesting that interesting insights about multivariate data can be obtained by looking at univariate quantiles of projections. Despite all the subtle differences aired here, the authors agree that HPS and KM have a lot in common: they both address depth contours by adopting a sort of directional approach. Regression aspects having been discussed to some extent in the previous section, they concentrate now on the possible impact on the computation of the depth contours (in the static situation, without covariates).
ISSN:0090-5364
2168-8966