Classification of mental task from EEG signals using Immune Feature Weighted Support Vector Machine

Abstract

The classification of mental tasks is one of the key issues of Brain Computer Interface (BCI). Owing to its powerful capacity in solving non-linearity problems, Support Vector Machine (SVM) has been widely used in classification. Traditional SVM, however, assumes that each feature of a sample contributes equally to classification accuracy, which is not necessarily true in real world applications. In addition, the parameters of SVM and kernel function also affect classification accuracy. In this paper, Immune Algorithm (IA) is introduced in searching for the optimal feature weights and parameters simultaneously. So Immune Feature Weighted SVM (IFWSVM) is used to multi-classify 5 kinds of mental tasks. Theoretical analysis and experimental results show that IFWSVM has better performance than Immune SVM (ISVM) without feature weight.