Classification Method Incorporating Decision Tree with Particle Swarm Optimization

This study attempts to develop a new method that could be used to handle the problem of finding the cut point or interval of continuous-valued attribute in decision tree, and to reach the following objectives: 1. The decision tree algorithm can handle the data that combines both the nominal attribute and continuous-valued attribute. 2. The decision tree algorithm has less nodes and branches in the situation that accuracy of prediction has no obvious change. 3. The decision tree algorithm can generate better rules with multi-interval division of attribute.