Decision Support System Using Decision Tree and Neural Networks

Abstract

Decision making in a complex and dynamically changing environment of the present day demands a new techniques of computational intelligence for building equally an adaptive, hybrid intelligent decision support system. In this paper, a Decision Tree-Neuro Based model was developed to handle loan granting decision support system and clinical decision support system(Eye Disease Diagnosis) which are two important decision problems that requires delicate care. The system uses an integration of Decision Tree and Artificial Neural Networks with a hybrid of Decision Tree algorithm and Multilayer Feed-forward Neural Network with backpropagation learning algorithm to build up the proposed model. Different representative cases of loan applications and eye disease diagnosis were considered based on the guidelines of different banks in Nigeria and according to patient complaint, symptoms and physical eye examinations to validate the model. Object-Oriented Analysis and Design (OO-AD) methodology was used in the development of the system, and an object-oriented programming language was used with a MATLAB engine to implement the models and classes designed in the system. The system developed, gives 88% success rate and eliminate the opacity of an ordinary neural networks system.

Keywords: Decision Tree-Neuro Based Model, Backpropagation Learning Algorithm, Object-Oriented Analysis and Design, MATLAB Embedded Engine, Loan Granting, Eye Diseases Diagnosis.