Publication

Abstract : Breast cancer is one of the deadly diseases amid women. The survival rate can be increased through early detection. The classification model with high level of predictive performance will help the medical experts to early identification of this disease. To develop such types of robust and optimal classification model, computational approach will be useful in early identification. In this paper, we introduce hybrid intelligent fuzzy-rough classification method based on rule induction. At initial stage, irrelevant features are removed through weak gamma evaluator. Performance of this classification model is examined for Wisconsin Breast Cancer Database (WBCD) and classification accuracy evaluated through F-measure. Performance measure of fuzzy-rough set optimization technique is taken into account by measuring sensitivity, specificity, and accuracy of the applied technique. Verification and validation exercise of the applied technique is carried out on the basis of results obtained in the similar direction by various realistic breast cancer images captured by thermography.