Breast cancer is one of the most common types of cancer that most often affect women. It is a leading cause of cancer death in less developed countries. So, it is important to characterize the severity of the disease as soon as possible. In this paper, we applied deep learning methods to determine the severity degree of patient with breast cancer using real data. The aim of the research is to characterize the severity of the disorder in a short time compared to the traditional methods. Deep learning methods are used because of their ability to detect target class more accurately than other machine learning methods, especially in the healthcare domain. In our research, several experiments were conducted using three different deep learning methods which are: Deep Neural Network (DNN), Recurrent Neural Network (RNN) and Deep Boltzmann Machines (DBM). Then, we compared their performance with the performance of traditional neural network method. We found that the f-measure of using the neural network is 74.52% compared to DNN which is 88.46 %, RNN which is 96.79% and DBM which is 97.28%.
Key words: Breast Cancer Severity, Medical data, deep learning, Deep Neural Network, Recurrent Neural Network and Deep Boltzmann Machines.
|
