Radial Basis Function Neural Network and Logistic Regression Analysis For Prognostic Classifi cation of Coronary Artery Disease Koroner Arter Hastalığının Sınıfl anmasında Radial Basis Fonsiyonu Sinir

Abstract

Objective: Artifi cial Neural Networks (ANNs) trained with backpropagation learning algorithm have been used commonly in previous studies. This study presents radial basis function neural network (RBFNN), a special kind of neural network, and logistic regression analysis (LRA) for prognostic classifi cation of Coronary Artery Disease (CAD). Methods:The records of 237 consecutive people who had been referred for the department of Cardiology were used in the analysis. Radial basis function neural network and logistic regression analysis were used for CAD classifi cation. Results:The results have shown that LRA and RBFNN were both successful for classifi cation and might be used for non-invasively based on clinical variables in the classifi cation of diseases like CAD. Conclusions:The work can be concluded that LRA performed the classifi cation better than RBFNN for prognostic CAD classifi cation in the present CAD data. However, RBFNN, utilizing larger sample sizes, can have better classifi cation accuracy. For more defi nite comparison, simulation studies should be carried out using various methods.