The Comparative Study of Deep Learning Neural Network Approaches for Breast Cancer Diagnosis

Abstract

Breast cancer is one of the life threatening cancer that leads to the most death due to cancer among the women. Early diagnosis might help to reduce mortality. Thus, this research aims to study on different approaches of the deep learning neural network model for breast cancer early detection for better prognosis. The performance of deep learning approaches such as Artificial Neural Network (ANN), Recurrent Neural Network (RNN) and Convolution Neural Network (CNN) are evaluated using the dataset from the University of Wisconsin. The findings show ANN achieved high accuracy of 99.9 % compared to others in detecting breast cancer. ANN is able to deliver better results with the provided dataset. However, more improvement needed for better performance to ensure that the approach used is reliable enough for breast cancer early diagnosis.