Study on the comparison of two different classifiers for EEG based non – invasive brain-computer interface: MLP neural networks vs. genetic algorithm

Journal of Bioengineering & Biomedical Science

ISSN: 2155-9538

Open Access

Study on the comparison of two different classifiers for EEG based non – invasive brain-computer interface: MLP neural networks vs. genetic algorithm

Nadiminti Krishna Teja, P S S Susheera, Sriram Narahari

: J Bioengineer & Biomedical Sci

Abstract :

BCI represents a direct communication between the neuron action of brain and a computer system. Electroencephalogram (EEG)-based Brain Computer Interfaces which have become a benchmark in the research of cognitive science, neuroscience, neural engineering etc. From the past 2 decades the field of BCI has a quick exponential growth rate and it also made a great interest in researches all over the world. Currently in BCI‐research the main focus is on people with severe motor disabilities. Indeed, in order to use a BCI, two phases are generally required: 1) an offline training phase which calibrates the system and 2) an online phase which uses the BCI to recognize mental states and translates them into commands for a computer. An online BCI requires following a closed-loop process, generally composed of six steps: brain activity measurement, preprocessing, feature extraction, classification, translation into a command and feedback. In the present study we have considered different parameters for the evaluation of the performances of the two classifiers in different parameters like accuracy, that is, percentage of correct classifications, training time and size of the training dataset. Here we adopted Multi Layer Perceptron Neural Network (MLP) and Genetic algorithm for classification of the feature vectors. The raw data which is collected from the dataset IIIa from the BCI III competition (BCI Competition III 2008) where two data files are available for each subject: training and testing. The results show that even if the accuracies of the two classifiers are quite similar, the MLP classifier needs a smaller training set to reach them with respect to Genetic algorithm used for classification. This leads to the preference of MLP for the classification Brain Computer Interface protocols.

Biography :

Boqiang Liu et all. Study of Feature Classification Methods in BCI Based on Neural Networks, Proceedings of the 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference Shanghai, China, September 1-4, 2005.K. Hara and K. Nakayamma, ?Comparison of activation functions in multilayer neural network for pattern classification? IEEE World Congress on Computational Intelligence., 1994, vol. 5, pp. 2997- 3002.C. Hernndez-Espinosa and M. Fernandez Redondo, ?Multilayer Feedforward Weight Initialization? European Symposium of Artificial Neural Networks 2001, pp. 119-124.

Google Scholar citation report
Citations: 289

Journal of Bioengineering & Biomedical Science received 289 citations as per Google Scholar report

Journal of Bioengineering & Biomedical Science peer review process verified at publons

Indexed In

arrow_upward arrow_upward