Analysis of Mental Task Ability in Students based on Electroencephalography Signals

The characterization of brain function during cognitive activity is gaining interest over the years. Electroencephalography (EEG) is extensively used in the medical field, cognitive science and brain-computer interface research. The brain’s electrical activity under cognitive load can be measured by electrodes placed on the scalp, and the observed signal is called EEG or brain wave. The cognitive load may be reading a book, listening to an online class, solving arithmetic problems or performing motor imagery functions. The ability of a person can be analyzed with the help of machine learning and signal processing techniques. In this paper, the EEG signals are taken from the Physionet database are analyzed. The signals were recorded from 36 students performing mathematical tasks. Different temporal, frequency and wavelet features are extracted using different signal processing techniques. Later, the features are classified using linear discriminant analysis (LDA), artificial neural networks (ANN), and random subspace classifier (RSC). The performance parameters obtained from the RSC classifier are 89.9123% accuracy, 99.4% precision, 92.8819% recall and 84.4664% F_score. The experimental outcome shows that the RSC classifier can better characterize brain activity under mental cognitive conditions.