TU Darmstadt / ULB / TUbiblio

Efficient prediction algorithms for binary decomposition techniques

Park, Sang-Hyeun and Fürnkranz, Johannes (2012):
Efficient prediction algorithms for binary decomposition techniques.
In: Data Mining and Knowledge Discovery, Springer Netherlands, pp. 40-77, 24, (1), ISSN 1384-5810, [Article]

Abstract

Binary decomposition methods transform multiclass learning problems into a series of two-class learning problems that can be solved with simpler learning algorithms. As the number of such binary learning problems often grows super-linearly with the number of classes, we need efficient methods for computing the predictions. In this article, we discuss an efficient algorithm that queries only a dynamically determined subset of the trained classifiers, but still predicts the same classes that would have been predicted if all classifiers had been queried. The algorithm is first derived for the simple case of pairwise classification, and then generalized to arbitrary pairwise decompositions of the learning problem in the form of ternary error-correcting output codes under a variety of different code designs and decoding strategies.

Item Type: Article
Erschienen: 2012
Creators: Park, Sang-Hyeun and Fürnkranz, Johannes
Title: Efficient prediction algorithms for binary decomposition techniques
Language: English
Abstract:

Binary decomposition methods transform multiclass learning problems into a series of two-class learning problems that can be solved with simpler learning algorithms. As the number of such binary learning problems often grows super-linearly with the number of classes, we need efficient methods for computing the predictions. In this article, we discuss an efficient algorithm that queries only a dynamically determined subset of the trained classifiers, but still predicts the same classes that would have been predicted if all classifiers had been queried. The algorithm is first derived for the simple case of pairwise classification, and then generalized to arbitrary pairwise decompositions of the learning problem in the form of ternary error-correcting output codes under a variety of different code designs and decoding strategies.

Journal or Publication Title: Data Mining and Knowledge Discovery
Volume: 24
Number: 1
Publisher: Springer Netherlands
Uncontrolled Keywords: aggregation, binary decomposition, efficient decoding, efficient voting, multiclass classification, pairwise classification, ternary ECOC
Divisions: 20 Department of Computer Science > Knowl­edge En­gi­neer­ing
20 Department of Computer Science
Date Deposited: 24 Jun 2011 13:31
Additional Information:

10.1007/s10618-011-0219-9

Identification Number: doi:10.1007/s10618-011-0219-9
Export:

Optionen (nur für Redakteure)

View Item View Item