Deep neural network with reduced feature for classification of breast cancer mammogram

Breast disease is the prevalent malignant growth in female all over the world and it is expanding in non-industrial nations, where most cases are analysed late. Mammography remains the best symptomatic advance from a treatment standpoint, despite widespread use and investigation of these images. The objective of this paper is to predict and classify the breast cancer using deep learning techniques. The extensive experiments are conducted on Wisconsin Demonstrative Bosom malignant growth (WDBC) dataset extricated from digitized pictures of Random MRI. Deep learning techniques such as deep neural network (DNN), recurrent neural network (RNN) and local linear radial basis function neural network (LLRBFNN) are used for experimental investigation. The performance of the proposed approach is experimented through various metrics such as accuracy, Jaccard index, precision, recall and F1 score.