A hybrid algorithm applied to classify medical datasets

In recent, the hybrid algorithm is one of the important approaches applied to classify medical datasets. In this paper, a new hybrid algorithm is proposed to classify medical datasets. In the proposed algorithm, scatter search is hybridized with support vector machine (SSHSVM). Furthermore, SSHSVM with feature selection (SSHSVMFS) is applied to boost classification accuracy and select significant features. Three medical datasets, colon, leukemia and lymphoma, were used to compare the performance of the proposed algorithm with other approaches. From experimental results, it shows that SSHSVMFS outperforms other existing approaches.