Discrimination of movement imagery EEG based on AR and SVM

Abstract

In the Brain-computer interface, classification and recognition technology plays an important role, especially the EEG classification and recognition for the movement imagery. In this paper, we use a new type of sensors to collect EEG signals. According to imagine the movement of left or right hand to identify two types of thinking, we proposed a new recognition method based on AR(auto-regressive) and SVM (support vector machine). In the identification process uses different kernel functions to classify comparison test. Compared to the traditional method based on support vector machine and Bayes, the correct rate has been greatly improved, and verifies the effectiveness of the method.