Neural-based decision trees classification techniques: a case study in water resources management

Abstract

This article compares the decision-tree algorithm (C4.5) and neural decision-tree algorithm (NDT) in the problem of water resources management. The feature of the NDT algorithm is the combination of the artificial neural network (ANN) technologies and the conventional decision-tree algorithm (C4.5) capabilities. The applicability of the presented algorithms is demonstrated through a case study of reservoir releases during typhoons. Shihmen Reservoir in Taiwan is the study site. The findings show superior performance of the NDT model in contrast to the traditional C4.5.