Improvement of decision tree ID3 algorithm

Abstract

This paper describes the basic concepts of the ID3 algorithm and its principles as well as the construction process. Because ID3 algorithm tends to select values for more attributes shortcomings, we introduce threshold, properties information gain rate and parameters to compensate for the lack of ID3 properties selected standard. Based on the above two points to achieve new property selection standard, the original ID3 algorithm is improved. Through the experiment, the improvements of the improved algorithm were compared. Experiment results show that the improved algorithm is effective.