To search, Click below search items.


All Published Papers Search Service


Comparing sensitivity of Radial Basis Function method with Multilayer Perceptron Network and Cox Proportional Hazard Model in Survival Data


Mohammad Salehi Veisi, Sadegh Rezaei and Khadejeh Salehivaysi


Vol. 17  No. 7  pp. 180-187


Cox proportional hazard model is broadly deployed in the analysis of survival data and reliability of systems and its application is contingent upon accepting some assumptions such as appropriateness of the risk. Neural network model, obviating the need of making any specific assumption, is an appropriate substitute in predicting survival data. To compare the sensitivity of neural network models with the Cox proportional hazard model, the present study investigates the sensitivity and specificity of radial basis function method, multilayer perceptron network, and cox proportional hazard model in the survival analysis of patients with myocardial infarction. The results of the study revealed that, compared to other models, neural network models performed better and were more precise in the survival analysis of patients with myocardial infarction. Moreover, compared to the other two methods, the radial basis function method is more sensitive, specific, and precise in the survival prediction of the patients under investigation in the present study and, accordingly, it is more reliable.


Radial Basis Function, network, Cox Proportional Hazard, model efficacy, survival analysis