Abstract
We propose a method to incrementally compute and use, during the concept formation, the relevance of a concept property. This relevance is computed through the account of the property correlation with other ones and it is used by the concept quality function in order to improve predictive accuracy. The proposed approach is analyzed concerning both the prediction power of the generated concepts and the time and space complexity of the concept formation algorithm
Chapter PDF
References
Decaestecker, C.: Apprentissage et outils statistiques en classification conceptuelle incrémentale. Revue d'Intelligence Artificielle, v 7, n. 1, 1993.
Fisher, D.H.: Knowledge Acquisition via Incremental Conceptual Learning. Machine Learning, vol 2, numero 2, 1987.
Fisher, D., Pazzani, M., Langley, P.: Concept Formation: Knowledge and Experience in Unsupervised Learning. Morgan Kaufmann, 1991.
Hanson, S., Bauer, M.: Conceptual Clustering, Categorization, and Polymorphy. Machine Learning, v. 3, pp: 343–372, 1989.
Reich, Y., Fenves, S.: The Formation and Use of Abstract Concepts in Design. In [Fisher 91].
Seifert, C.: A Retrieval Model Using Feature Selection. Proc. of the 6th Int. Workshop on ML. Morgan Kaufmann. 1989.
Stepp, R., Michalski, R.: Conceptual Clustering: Inventing goal oriented classifications of structured objects. In Michalski, R., Carbonnel, J., Mitchell, T. (Eds), Machine Learning, An Intelligence Approach. Vol II. Morgan Kaufmann, CA. 1986.
Vasco, J.J.F., Faucher, C., Chouraqui, E.: A Knowledge Acquisition Tool for Multi perspective Concept Formation. In proc. of European Knowledge Acquisition Workshop — EKAW-96. Shadbolt, Shreiber and O'Hara (Eds). Springer Verlag; 1996.
Vasco, J.J.F., Formation de concepts dans le contexte des langages de schémas. Thèse de doctorat. Université d'Aix Marseille III, IUSPIM/DIAM, 1997.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1998 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Furtado Vasco, J.J. (1998). Determining property relevance in concept formation by computing correlation between properties. In: Nédellec, C., Rouveirol, C. (eds) Machine Learning: ECML-98. ECML 1998. Lecture Notes in Computer Science, vol 1398. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0026701
Download citation
DOI: https://doi.org/10.1007/BFb0026701
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-64417-0
Online ISBN: 978-3-540-69781-7
eBook Packages: Springer Book Archive