| Title |
An Optimized Deep Belief Network Model for Accurate Breast Cancer Classification |
| Authors |
(Bunyodbek Ibrokhimov) ; (Cheonghwan Hur) ; (Hyunseok Kim) ; (Sanggil Kang) |
| DOI |
https://doi.org/10.5573/IEIESPC.2020.9.4.266 |
| Keywords |
Breast cancer; Deep belief network; Particle swarm optimization; Classification; Feature selection |
| Abstract |
Breast cancer is the second most common cause of mortality in women. In the last decade, the rate of new diagnoses has increased significantly. Meanwhile, according to the World Health Organization, breast cancer can be treated effectively if it is detected in the early stages. Computer automated diagnosis systems have greatly helped researchers to learn the evolution and causes of breast cancer, as well as to monitor and detect the disease among women. However, there is a need for modern and accurate models to reduce the risk of breast cancer as well as for further improvements in diagnostic performance. In this paper, a deep belief network (DBN) model is proposed to achieve high accuracy in breast cancer classification. A particle swarm optimization algorithm is employed and integrated into the DBN to optimize the model’s parameters. The model is tested on two popular breast cancer datasets. Moreover, analysis of the datasets and improvements in data quality using data mining are shown, and an effective feature-selection method for a high-dimensional feature space is discussed. In experiments, the proposed method showed robust performance with high classification accuracy, matching, and even outperforming existing state-of-the-art algorithms. |