DA-Based Parameter Optimization of Combined Kernel Support Vector Machine for Cancer Diagnosis

摘要: As is well known, the correct diagnosis for cancer critical to save patients’ lives. Support vector machine (SVM) has already made an important contribution field of classification. However, different kernel function configurations and their parameters will significantly affect performance SVM classifier. To improve classification accuracy classifier diagnosis, this paper proposed a novel algorithm based on dragonfly with combined (DA-CKSVM) which was constructed from radial basis (RBF) polynomial kernel. Experiments were performed six data sets University California, Irvine (UCI) learning repository two Cancer Program Legacy Publication Resources evaluate validity algorithm. Compared four well-known algorithms: algorithm-SVM (DA-SVM), particle swarm optimization-SVM (PSO-SVM), bat (BA-SVM), genetic (GA-SVM), able find optimal achieved better datasets.