A hybrid decision tree/genetic algorithm for coping with the problem of small disjuncts in data mining

Abstract

The problem of small disjuncts is a serious challenge for data mining algorithms. In essence, small disjuncts are rules covering a small number of examples. Due to their nature, small disjuncts tend to be error prone and contribute to a decrease in predictive accuracy. This paper proposes a hybrid decision tree/genetic algorithm method to cope with the problem of small disjuncts. The basic idea is that examples belonging to large disjuncts are classified by rules produced by a decision-tree algorithm, while examples belonging to small disjuncts (whose classification is considerably more difficult) are classified by rules produced by a genetic algorithm specifically designed for this task. 1 INTRODUCTION In the context of the well-known classification task of data mining, the discovered knowledge is often expressed as a set of IF-THEN rules, since this kind of knowledge representation is intuitive for the user. From a logical viewpoint, typically the discovered rules ar...