Feature Construction and δ-Free Sets in 0/1 Samples

Abstract : Given the recent breakthrough in constraint-based mining of local patterns, we decided to investigate its impact on feature construction for classification tasks. We discuss preliminary results concerning the use of the so-called δ-free sets. Our guess is that their minimality might help to collect important features. Once these sets are computed, we propose to select the essential ones w.r.t. class separation and generalization as new features. Our experiments have given encouraging results.
Type de document :
Communication dans un congrès
Discovery Science, 9th International Conference, DS 2006, Oct 2016, Barcelona, Spain. Springer, Proceedings of Discovery Science, 9th International Conference, 4265, pp.363-367, 2006, Lecture Notes in Computer Science, LNCS
Liste complète des métadonnées

http://hal.univ-reunion.fr/hal-01464502
Contributeur : Réunion Univ <>
Soumis le : vendredi 10 février 2017 - 11:43:23
Dernière modification le : mercredi 31 octobre 2018 - 12:24:25

Identifiants

  • HAL Id : hal-01464502, version 1

Citation

Nazha Selmaoui-Folcher, Claire Leschi, Dominique Gay, Jean-François Boulicaut. Feature Construction and δ-Free Sets in 0/1 Samples. Discovery Science, 9th International Conference, DS 2006, Oct 2016, Barcelona, Spain. Springer, Proceedings of Discovery Science, 9th International Conference, 4265, pp.363-367, 2006, Lecture Notes in Computer Science, LNCS. 〈hal-01464502〉

Partager

Métriques

Consultations de la notice

231