Decision Tree Algorithm Considering Distances Between Classes

Decision tree algorithm (DT) is a commonly used data mining method for classification and regression. DT repeatedly divides dataset into pure subsets based on impurity measurements such as entropy Gini. Then relatively “pure” partitions consisting of observations with the (almost) same class are obtained. Gini index one representative indices measuring data. However, does not take account distances between classes. If classes considered when impurity, decision can distinguish clearly different To end, new Rao-Stirling proposed considering considers in way that weights more to pairs references distant impurity. Experimental results indicate superior terms accuracy, implying help improve accuracy DT.