A Novel Decision Tree Algorithm
Weili Jiao, Mingxin Zhang · International Conference on E-Business and E-Government · 2012
In the process of constructing a decision tree, the criteria of selecting splitting attributes will influence the efficiency of classification. We present Class division degree decision tree method, a new concept based on Category factor which explains the classification ability of attributes set on the sample space. We have made further improvement to the algorithm and propose the concept of an optimized coefficient, this algorithm is used to choose the test property in the decision tree construction. The experiment shows that the algorithm has better efficiency of building the decision tree compared with the WMR algorithm based approach.