Refining Invariant Coordinate Selection via Local Projection Pursuit

Dümbgen, Lutz; Gysel, Katrin; Perler, Fabrice (2023). Refining Invariant Coordinate Selection via Local Projection Pursuit. In: Yi, Mengxi; Nordhausen, Klaus (eds.) Robust and Multivariate Statistical Methods: Festschrift in Honor of David E. Tyler (pp. 121-136). Cham: Springer International Publishing 10.1007/978-3-031-22687-8_6

[img] Text
LocalProjPursuit.pdf - Accepted Version
Restricted to registered users only
Available under License Publisher holds Copyright.

Download (565kB)
[img] Text
978-3-031-22687-8_6.pdf - Published Version
Restricted to registered users only
Available under License Publisher holds Copyright.
Author holds Copyright

Download (609kB)

Invariant coordinate selection (ICS), introduced by Tyler et al. (J. Roy. Stat. Soc. B 71(3):549–592, 2009), is a powerful tool to find potentially interesting projections of multivariate data. In some cases, some of the projections proposed by ICS come close to really interesting ones, but little deviations can result in a blurred view which does not reveal the feature (e.g., a clustering), which would otherwise be clearly visible. To remedy this problem, we propose an automated and localized version of projection pursuit (PP), cf. Huber (Ann. Stat. 13(2):435–525, 1985). Precisely, our local search is based on gradient descent applied to estimated differential entropy as a function of the projection matrix.

Item Type:

Book Section (Book Chapter)

Division/Institute:

08 Faculty of Science > Department of Mathematics and Statistics > Institute of Mathematical Statistics and Actuarial Science

UniBE Contributor:

Dümbgen, Lutz

Subjects:

300 Social sciences, sociology & anthropology > 360 Social problems & social services
500 Science > 510 Mathematics

ISBN:

978-3-031-22686-1

Publisher:

Springer International Publishing

Language:

English

Submitter:

Lutz Dümbgen

Date Deposited:

25 Apr 2023 12:27

Last Modified:

25 Apr 2023 12:35

Publisher DOI:

10.1007/978-3-031-22687-8_6

BORIS DOI:

10.48350/181950

URI:

https://boris.unibe.ch/id/eprint/181950

Actions (login required)

Edit item Edit item
Provide Feedback