Diagnosis of breast cancer with an innovative adaptive Support Vector Machine

In this study, a novel methodology based on Support Vector Machine (SVM) is proposed. In the proposed method, the sigma value belonging to the radial based function which is being used as the kernel function for the support vector machine is computed by using an adaptive mechanism. By this means, a new kind of SVM which can be defined as “Adaptive SVM” (ASVM) is proposed, and smart diagnosis of the breast cancer is aimed. During the training and test phases of this newly designed smart system, the prognostic breast cancer dataset which is provided from University of California is used. It is observed that the novel methodology which is firstly proposed in this study has a correct classification rate of 94.29% on the prognostic breast cancer dataset.