Motor imagery classification based on the optimized SVM and BPNN by GA

Brain-computer interface (BCI) is a specific Human-Computer interface in which the brain wave is employed as the carrier of control information. The ultimate goal of BCI is to build a direct communication pathway between human brain and external environment that does not depend on the limb mobility and language. In this paper, we carry out the experiment about the left or right hand motor imagery, and support vector machine with genetic algorithm (GA-SVM) and back propagation neural network with genetic algorithm (GA-BP) are employed to classify the μ rhythm evoked by movement imagination. The experiment results prove that GA-SVM can easily find out the appropriate parameters of SVM and GA-BP can avoid getting into local minimization to great extend. So higher accuracy of classification is achieved.