Feature transformation methods in data mining

Abstract

The quality of knowledge extracted from a data set can be enhanced by its transformation. Discretization and filling missing data are the most common forms of data transformation. A new transformation method named feature bundling is introduced. A feature bundle involves a set of features in its pure or transformed form. The computational results reported in this paper show that the classification accuracy of decision rules generated from data sets with feature bundles is enhanced. The proposed concept of feature bundling is applied to a data set from the semiconductor industry.