A New Hybrid Method of Generation of Decision Rules Using the Constructive Induction Mechanism

摘要

Our research is devoted to develop a new method of generation of a set of decision rules. This method is compiled using two different mechanisms. The first one is based on applying a new constructive induction algorithm to the investigated dataset. The belief networks are used in this algorithm. The aim is to find the most important descriptive attribute that is calculated on the basis of other attributes. The second part of the presented method constitutes the improvement algorithm that is used in an optimization process of a gathered rule set. The results of our research contain the comparison of classification efficiency using several datasets.