A Retractile Decision Tree Algorithm Based on Influence Degree
Zhonghua Li · Journal of Hebei North University · 2008
Objective Aiming at high complex degree of the algorithm ID3,to introduce an advanced algorithm,DTA(Decision Tree Algorithm),which is based on decision tree.Methods Using influence degree as the standard of feature selection for a good calculation complex degree and using a new data structure which is an attribute list based on classification for a good retractility.Results As is shown from the result of experiment,the advanced algorithm can make accurate decision tree and better complex degree than traditional algorithms.Conclusion The algorithm can generate correct decision trees and obtain relevant decision rules on the conditions of low configuration of computer and less consumption of resourses.