TU Darmstadt / ULB / TUbiblio

Efficient Discovery of Expressive Multi-label Rules Using Relaxed Pruning

Klein, Yannik ; Rapp, Michael ; Loza Mencía, Eneldo
Kralj Novak, Petra ; Šmuc, Tomislav ; Džeroski, Sašo (eds.) (2019):
Efficient Discovery of Expressive Multi-label Rules Using Relaxed Pruning.
pp. 367-382, DS'19 - 22nd International Conference on Discovery Science, Split, Croatia, October 28.–30., 2019, ISBN 978-3-030-33778-0,
DOI: 10.1007/978-3-030-33778-0_28,
[Conference or Workshop Item]

Abstract

Being able to model correlations between labels is considered crucial in multi-label classification. Rule-based models enable to expose such dependencies, e.g., implications, subsumptions, or exclusions, in an interpretable and human-comprehensible manner. Albeit the number of possible label combinations increases exponentially with the number of available labels, it has been shown that rules with multiple labels in their heads, which are a natural form to model local label dependencies, can be induced efficiently by exploiting certain properties of rule evaluation measures and pruning the label search space accordingly. However, experiments have revealed that multi-label heads are unlikely to be learned by existing methods due to their restrictiveness. To overcome this limitation, we propose a plug-in approach that relaxes the search space pruning used by existing methods in order to introduce a bias towards larger multi-label heads resulting in more expressive rules. We further demonstrate the effectiveness of our approach empirically and show that it does not come with drawbacks in terms of training time or predictive performance.

Item Type: Conference or Workshop Item
Erschienen: 2019
Editors: Kralj Novak, Petra ; Šmuc, Tomislav ; Džeroski, Sašo
Creators: Klein, Yannik ; Rapp, Michael ; Loza Mencía, Eneldo
Title: Efficient Discovery of Expressive Multi-label Rules Using Relaxed Pruning
Language: English
Abstract:

Being able to model correlations between labels is considered crucial in multi-label classification. Rule-based models enable to expose such dependencies, e.g., implications, subsumptions, or exclusions, in an interpretable and human-comprehensible manner. Albeit the number of possible label combinations increases exponentially with the number of available labels, it has been shown that rules with multiple labels in their heads, which are a natural form to model local label dependencies, can be induced efficiently by exploiting certain properties of rule evaluation measures and pruning the label search space accordingly. However, experiments have revealed that multi-label heads are unlikely to be learned by existing methods due to their restrictiveness. To overcome this limitation, we propose a plug-in approach that relaxes the search space pruning used by existing methods in order to introduce a bias towards larger multi-label heads resulting in more expressive rules. We further demonstrate the effectiveness of our approach empirically and show that it does not come with drawbacks in terms of training time or predictive performance.

ISBN: 978-3-030-33778-0
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Knowl­edge En­gi­neer­ing
Event Title: DS'19 - 22nd International Conference on Discovery Science
Event Location: Split, Croatia
Event Dates: October 28.–30., 2019
Date Deposited: 18 Dec 2019 10:38
DOI: 10.1007/978-3-030-33778-0_28
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details