Abstract: Design of Clinical Decision Support System (ICDSS) is a challenging task in which a set of symptoms is possible to have distinct diseases and each disease has the possibility to have different categories or labels, therefore the usage of Multi-Label Classification (MLC) is required in the DD. MLC refers to the problem where each instance is associated with more than one class labels. Multi-Label Data (MLD) are high dimensional and deteriorates the performance of the classifier in terms of diagnostic accuracy. Classification of MLD is a very challenging task by existing methods and need a systematic approach. In this study, an efficient feature selection with ensemble Deep Learning (DL) algorithm for handling MLC problems is proposed. The effectiveness of the proposed Multi-Label Ensemble Deep Learning (MED) algorithm is investigated with two publicly available ML medical data using various evaluation measures. The MED results significant improvement in the performance compared with existing methods in the literature. The results reveal some interesting conclusion with respect to the use of the proposed approach to help the medical practitioners in a better decision making in the diagnosis and treatment with the least number of symptoms in the MLD.
D. Senthilkumar and S. Paulraj, 2016. Ensemble Deep Learning for Multi Label Classification in the Design of Clinical Decision Support System. Asian Journal of Information Technology, 15: 2632-2637.