An information entropy based splitting criterion better for the Data Mining Decision Tree algorithms
Laviniu Aurelian Badulescu · 2018
This paper attempts to discover a better entropy-based splitting criterion for the induction of the Decision Trees (DT) than the entropy-based splitting criteria known so far. Eight entropy-based splitting criteria along with the performance tests carried out by three DT built on their basis, were taken into account: unpruned DT, pessimistically pruned DT and error-based pruned DT. All of these DT types were trained on seven databases, and then executed on their test data. Our experiments highlight the very good performances achieved by the square information gain ratio1 splitting criterion, which has shown always a better behavior than the information gain ratio criterion used in C4.5 algorithm and than other six information entropy based splitting criteria.