Discretization is a kind of preprocessing of data in which numerical attributes are converted into symbolic attributes. A discretization algorithm, based on hierarchical cluster analysis, was implemented by Michal R. Chmielewski. The idea of globalization of local discretization algorithms was tested experimentally by Michal R. Chmielewski as well. The entire family of discretization algorithms was implemented by Xiaoqiang Yao. Another approach, based on including discretization to rule induction process was experimentally tested by Anitha Lakshmanan. Yet another algorithm, in which intervals were merged together after rule induction, was implemented and tested by Munish Malik.
Return to the Main Page