Fuzzy radial basis function neural networks for web applications cost estimation

The Fuzzy Radial basis function Neural Networks (FRBFN) for software cost estimation is designed by integrating the principles of RBFN and the fuzzy C-means clustering algorithm. The architecture of the network is suitably modified at the hidden layer to realise a novel neural implementation of the fuzzy clustering algorithm. Fuzzy set-theoretic concepts are incorporated at the hidden layer, enabling the model to handle uncertain and imprecise data, which can improve greatly the accuracy of obtained estimates. MMRE and Pred are used as measures of prediction accuracy for this comparative study. The results show that an RBFN using fuzzy C-means performs better than an RBFN using hard C-means. This study uses data on web applications from the Tukutuku database.