A performance analysis of user's intention classification from EEG signal by a computational intelligence in BCI

Abstract

Knowing the user's intentions is very important and can be useful in our daily life. It would be a very useful way, especially if people with disabilities can use these functions as a means of self-expression. The user's intension classification is a kind of common time-series problem for detecting human cognitive state. In this paper, we classify user intention by analyzing EEG signal using machine learning in BCI. The performance of the classification accuracy can be achieved by using the proposed approach in terms of the number of neurons in the hidden layer, which also leads types of membership function in fuzzy rules. We prepared training and test data using the Emotive headset for the experiment. Our experimental results show that the proposed approach gives us a quite promising method with 5 fuzzy rules obtained through a fuzzy C-means clustering. It is a simple fuzzy system with neural network structure by tuning GA providing statistically superior solutions. Experimental results show that the best results were obtained using the electrode position {F7, F8, FC5, FC6} of EEG. Experimental results using training data showed an accuracy of 94.2%. However, the result of using the test data after learning shows a slightly lower accuracy of 92.3%. This experiment shows that using training data and test dares can result in more than 90% accuracy. Experimental results show that all 4--action behaviors have similar accuracy.