A Study on Brain Computer Interface using Learning Vector Quantization

Brain-computer interface (BCI) can provide communication and control between the human brain and a computer. Detection of brain signals is the most basic level for these systems. Magnetoencephalography (MEG) is a non-invasive neuroimaging technique for decoding brain activity. MEG signals are complicated and can be easily affected by environmental events and functional differences of the brain. It is difficult to get information from these complex signals for BCI systems. Therefore, advanced signal processing techniques are required to make the information meaningful. In this study, the success of learning vector quantization (LVQ) algorithm has been put forward by classifying magnetoencephalography signals through LVQ. Classification accuracy is obtained via 10-fold cross validation. The performance of proposed classifier is compared with the results of the previous methods reported focusing on MEG and using same dataset.