Design of an early prediction approach BC-DEF for breast cancer using Ensemble Learning for Fisher Linear Discriminant feature selection approach

Abstract

This research work (BC-DEF ) can be presented as a breast cancer prediction framework using diagnostic mammogram images and a clinical dataset based on biopsy. The proposed solution aims to provide an explainable classification model using strongly dependent quality of the extracted mammographic features of breast cancer affected patients. The classification approach uses BC-DEF as proposed framework which defines an Ensemble Learning for Fisher Linear Discriminant for Breast Cancer prediction. The proposed work BC-DEF focuses on improved performance metrics compared to traditional classifiers based breast cancer tumour pattern recognition algorithms. Observed accuracy is found to be 98.7 % better than other classifier models.