Proposal for Measuring Quality of Decision Trees Partition

Decision Trees

Decision trees methods are widely used in the classification field; they provide good results as far as classification techniques are concerned, but involve complexity in the formulas used for calculations. They enable building classification models as the latter closely resemble human reasoning and are easy to understand. To have more information about decision trees methods, the reader can refer to (Kotsiantis, 2013); the paper describes basic decision trees issues and current research points. In the chapter published by (Barros, 2015) are presented in details the most common approaches for decision-tree induction (Top Down Induction) with brief comments on some alternative construction methods summarizing the main design options to building decision-tree induction algorithms.