DefinePK

Abstract

Breast cancer is among the commonest cancer in the world, as the incidence rates have been on the rise and have posed a big burden to the healthcare systems. To increase the chance of survival and decrease the mortality, it is important to detect and diagnose the problems early and properly. Recent literature has applied machine learning and deep learning to the classification of breast cancer, but numerous models, such as Support Vector Machines (SVM) and Random Forests, have been challenged by lack of accuracy and generalization. We used a Deep Neural Network (DNN) to determine benign and malignant tumors using the Wisconsin Diagnostic Breast Cancer (WDBC) dataset in this study. The data has five hundred six hundred and ninety-nine samples and thirty features which were obtained as nuclei of breast mass cells, the data has undergone pre-processing by eliminating irrelevant data and has been standardized to achieve a better model performance.  In order to avoid overfitting, the DNN architecture was trained until 50 epochs and early stopping was implemented after 36 epochs.  The model was successful at 98.63 and 97.83 in training and validation accuracy respectively, and it can be inferred that the model is performing well in the classification of breast tumors. The findings suggest that the developed DNN model is a viable instrument in the early diagnosis of breast cancer, which is a crucial issue in the prior machine learning models