In classification tasks it may sometimes not be meaningful to build single rules on the whole data. This may especially be the case if the classes are composed of several subclasses. Several common as well as recent issues are presented to solve this problem. As it can e.g. be seen in Weihs et al. (2006) there may result strong benefit from such local modelling. All presented methods are evaluated and compared on four real-world classification problems in order to obtain some overall ranking of their performance following an idea of Hornik and Meyer (2007).
|Title of host publication
|Advances in Data Mining: Medical Applications, E-Commerce, Marketing, and Theoretical Aspects
|Subtitle of host publication
|Proceedings of 8th Industrial Conference, ICDM 2008
|Number of pages
|Published - 2008
|Lecture Notes in Computer Science