Object selection based on subclass error correcting for ALVOT

Abstract

ALVOT is a supervised classification model based on partial precedences. These classifiers work with databases having objects described simultaneously by numeric and nonnumeric features. In this paper a new object selection method based on the error per subclass is proposed for improving the accuracy, especially with noisy training matrixes. A comparative numerical experiment was performed with different methods of object selection. The experimental results show a good performance of the proposed method with respect to previously reported in the literature. © Springer-Verlag Berlin Heidelberg 2007.