Publication

Abstract : Breast cancer is currently one of the deadliest types of cancer, and the death rate has considerably grown as a result of a lack of knowledge about the disease, its symptoms, and prevention techniques. Therefore, early detection at a nearly stage is crucial and vital in order to limit the progression of cancer. Malignant and benign breast cancer are the two further subtypes. In order to categorise the different forms of breast cancer, an automated system with logistic regression and neural network is proposed. The classification of the breast cancer data utilises the DNN with various levels of processing. The Digital Database for Screening Mammography (DDMS) dataset was used in the proposed study on the Kaggle platform. The dataset was divided into various train-test splits. On the basis of accuracy, ROC AUC cuve the system’s performance is evaluated. The outcome demonstrates that proposed-2 outperforms in a comparable sense with a training accuracy of 98.99% and testing accuracy of 98.83%.