Universität zu Köln

Mozharovskyi, Pavlo, Mosler, Karl and Lange, Tatjana (2015). Classifying real-world data with the DD alpha-procedure. Adv. Data Anal. Classif., 9 (3). S. 287 - 315. HEIDELBERG: SPRINGER HEIDELBERG. ISSN 1862-5355

Full text not available from this repository.

Abstract

The -classifier, a nonparametric fast and very robust procedure, is described and applied to fifty classification problems regarding a broad spectrum of real-world data. The procedure first transforms the data from their original property space into a depth space, which is a low-dimensional unit cube, and then separates them by a projective invariant procedure, called -procedure. To each data point the transformation assigns its depth values with respect to the given classes. Several alternative depth notions (spatial depth, Mahalanobis depth, projection depth, and Tukey depth, the latter two being approximated by univariate projections) are used in the procedure, and compared regarding their average error rates. With the Tukey depth, which fits the distributions' shape best and is most robust, 'outsiders', that is data points having zero depth in all classes, appear. They need an additional treatment for classification. Evidence is also given about the dimension of the extended feature space needed for linear separation. The -procedure is available as an R-package.

Item Type:	Journal Article
Creators:	Creators Email ORCID ORCID Put Code Mozharovskyi, Pavlo UNSPECIFIED UNSPECIFIED UNSPECIFIED Mosler, Karl UNSPECIFIED UNSPECIFIED UNSPECIFIED Lange, Tatjana UNSPECIFIED UNSPECIFIED UNSPECIFIED
URN:	urn:nbn:de:hbz:38-394504
DOI:	10.1007/s11634-014-0180-8
Journal or Publication Title:	Adv. Data Anal. Classif.
Volume:	9
Number:	3
Page Range:	S. 287 - 315
Date:	2015
Publisher:	SPRINGER HEIDELBERG
Place of Publication:	HEIDELBERG
ISSN:	1862-5355
Language:	English
Faculty:	Unspecified
Divisions:	Unspecified
Subjects:	no entry
Uncontrolled Keywords:	Keywords Language DATA DEPTH; CLASSIFICATION; REGRESSION Multiple languages Statistics & Probability Multiple languages
URI:	http://kups.ub.uni-koeln.de/id/eprint/39450