BCI based EEG Signals classification using Machine Learning Approach

this research investigates the effectiveness of classifiers in non-invasive electroencephalogram (EEG) Classification based brain computer interface (BCI) systems. Such systems are crucial for improving the quality of life for people with neuromuscular disorders. The study uses EEG data from 25 healthy individuals over three experimental sessions to evaluate the effectiveness of k-nearest neighbor (KNN) and Multi-layer Perceptron (MLP) classifiers, as well as support vector machine (SVM) and random forest (RF) classifiers. The data were processed using signal augmentation techniques, such as a 4th band-pass Butterworth filter. Then, the data underwent evaluation through the application of 12-time domain feature extraction techniques. The results show that SVM performs better than RF with a 98.75% classification accuracy. It proves its superior ability to distinguish between left-and right-hand movement patterns. Similarly, MLP outperforms KNN with an accuracy of 95%, highlighting its potential for creating reliable online BCI applications. These findings pave the way for future developments in Brain Computer Interfacing (BCI) technology, which aims to provide dependable and effective user interfaces for people with motor disabilities.