Quantitative Analysis of EEG Neurofeedback using optimized 1-DPSO

Brain-Computer Interface (BCI) systems are the leading technology in the world related to Neurosciences. Human intelligence and imagination have no bounds and this has led to vast advancement in BCI systems related to Neurological and allied sciences. This system measures the activity of the Central Nervous System (CNS) using biosignals and output is called Electroencephalography (EEG). The main purpose of BCI is to acquire EEG brain signals, identify patterns, extract features, and produce resultant actions. This process communicates with a modest electronic system designed for movements of physically challenged or paralyzed people. The purpose of this BCI system is to design a model to check the attention level of body movement. The movements are based on the EEG signals captured from the 19-electrode EEG headset. This allows gaining control over optimized real-time feature selection for EEG signals. The dataset of 30 subjects’ sample EEG signals is recorded for classification and analysis purpose. The EEG signals are classified using Logistic Regression, Decision Tree, and Random Forest algorithms. It is shown that Random Forest is the most efficient classifier with the highest accuracy of 99.47%.