Dhoriva Urwatul Wutsqa, Agus Maman Abadi, Nurhayadi
In this study, breast cancer has been successfully classified by using a hybrid model of fuzzy and neural network. For this purpose, we propose two-hybrid models namely fuzzy neural network (fuzzy NN) and fuzzy radial basis function neural network (fuzzy RBFNN). The backpropagation algorithm is employed to estimate the weights of the fuzzy NN model. The K-Means clustering and singular value decomposition are developed to estimate the parameters of the fuzzy RBFNN. The benign and malignant breast tissues data drawn from Wisconsin Breast Cancer Database (WBCD) and Wisconsin Diagnostic Breast Cancer (WDBC) are used in the classification. The variables in the data sets are the features from the digitized images of fine needle aspiration (FNA) biopsy of the breast. The result shows that both models deliver high accuracies on WBCD and WDBC data sets. However, the fuzzy NN shows slightly better performance than the fuzzy RBFNN © 2022. IAENG International Journal of Computer Science.All Rights Reserved.
Statistics Department, Universitas Negeri Yogyakarta, Indonesia; Mathematics Department, Universitas Negeri Yogyakarta, Indonesia; Mathematics Education Department, Tadulako University, Indonesia
