A fuzzy measure of similarity for instance-based learning
Instance-based learning techniques are based on computing distances or similarities between instances with known classification and objects to classify. The nearest instance or instances are used to predict the class of unseen objects. In this paper we present a fuzzy measure of similarity between fuzzy sets and between elements. This measure allows us to obtain a normalized value of the proximity of objects defined by fuzzy features.
In order to test the efficiency of the proposed measure, we use it in a simple instance-based learning system and make a comparison with other measures proposed in literature.
Unable to display preview. Download preview PDF.
- 2.Batchelor, B.G.: Pattern Recognition: Ideas in Practice. Plenum Press, New York (1978)Google Scholar
- 4.Blake, C., Keogh, E., Merz, C. J. UCI Repository of machine learning databases http://www.ics.uci.edu/mlearn/MLRepository.html. Irvine, CA: University of California.Google Scholar
- 5.Botana, F.: Deriving fuzzy subsethood measures from violations of the implication between elements. Lec. Not. Artif. Intel. 1415 (1998) 234–243Google Scholar
- 7.Diday, E.: Recent progress in distance and similarity measures in pattern recognition. II Int. Joint Conf. on Pattern Recognition (1974) 534–539Google Scholar
- 9.Dubois D., Prade, H.: Fuzzy Sets and Systems: Theory and Applications. Academic Press, New York (1980)Google Scholar
- 13.Quinlan, J.R.: C4.5. Programs for Machine Learning. Morgan Kaufmann, San Mateo (1993)Google Scholar