A 3-Layer Deep Learning Model for Predicting Breast Cancer Disease

Abstract

Breast cancer is the major form of cancer attacking Malaysian women. Almost one over 20 women in Malaysia has been diagnosed to have this disease. Breast cancer is also called as malignant (cancerous) tumor initiating from the cells of the breast. Breast cancer disease is shown to contribute a high rate of mortality for women in worldwide. Thus, a mechanism for an early detection and prediction is of a great concern. The Machine Learning (ML) and Deep Learning (DL) approaches are shown to have a significant impact on the process of breast cancer prediction for an early awareness. Both machine learning and deep learning have become a research hotspot due to the highest prediction accuracy percentage have been reported. As for deep learning, it derives from a machine learning technique, which is neural network. This research attempts to explore on deep neural network or also called deepnets. How well the neural network will be used on how many layers of the neural network can contribute to the highest percentage of prediction accuracy. This paper presents the results of breast cancer prediction prior to adopting a 3-layer deep learning approach. We went through a 3-layer deep neural network in this research on the Wisconsin Breast Cancer Diagnostic (WBCD) dataset. The results of the experimentation reveal the significant prediction accuracy of approximately 0.9809 as the indication of our proposed deep learning approach has given a significant contribution in breast cancer disease.