Support vector machine based ensemble classifier

The strategy that the original input space is partitioned into several input subspaces usually works for improving the performance. Different from conventional partition methods, the partition method, attribute reduction based on rough sets theory, allows the input subspaces partially overlapped. These input subspaces can offer complementary information about hidden data patterns. In every subspace, a SVM sub-classifier is learned. Then, those SVM sub-classifiers with good performance are selected and combined to construct an ensemble classifier. The proposed method is applied to decision-making of medical diagnosis. Comparison between our method and several other popular ensemble methods is done. Experimental results demonstrate that the proposed approach can make full use of the information contained in data and improve the decision-making performance.