Title :
Supervised nonparametric information theoretic classification
Author :
Archambeau, Cédric ; Butz, Torsten ; Popovici, Vlad ; Verleysen, Michel ; Thiran, Jean-Philippe
Author_Institution :
Machine Learning Group, Univ. Catholique de Louvain, Belgium
Abstract :
In this paper, supervised nonparametric information theoretic classification (ITC) is introduced. Its principle relies on the likelihood of a data sample of transmitting its class label to data points in its vicinity. ITC´s learning rule is linked to the concept of information potential and the approach is validated on Ripley´s data set. We show that ITC may outperform classical classification algorithms, such as probabilistic neural networks and support vector machines.
Keywords :
information theory; learning (artificial intelligence); maximum likelihood estimation; neural nets; nonparametric statistics; pattern classification; support vector machines; Ripley data set; information theoretic classification; learning rule; probabilistic neural network; supervised nonparametric classification; support vector machines; Classification algorithms; Clustering algorithms; Machine learning; Neural networks; Pattern recognition; Prototypes; Shape; Signal processing algorithms; Support vector machine classification; Support vector machines;
Conference_Titel :
Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Conference on
Print_ISBN :
0-7695-2128-2
DOI :
10.1109/ICPR.2004.1334554
