Mental State Prediction by Deployment of Trained SVM Model on EEG Brain Signal

Abstract

Mental States are a function of brain activity; with advancements in Brain Computer Interface (BCI) tools, they can be effectively predicted. Generally, BCI researches are sophisticated requiring multi-channel electrodes, and often carried out in controlled lab environment. This paper illustrates that a simple BCI research, targeting specific region of brain, can be conducted with an elementary setup. A method is demonstrated to predict the level of attentiveness using Electroencephalography (EEG) signal obtained from single-channel, non-invasive, dry-electrode placed on prefrontal cortex region. The acquired raw signal was processed to obtain features which were used to classify the attention state into three classes. As Support Vector Machine (SVM) is more effective in training and classification of high dimensional data, it was implemented and its results were studied.