Abstract
Emotions play an important role in our thinking and behavior and hence contribute in shaping up of our personality. Many theoretical and experimental researches have been conducted to recognize the emotions from verbal or non-verbal behaviors. It is well known that the electroencephalogram (EEG) signals contain rich information about the activities of the brain and they can reliably enable us to estimate the emotions if they are properly interpreted. In this paper, we propose a model to discriminate the emotional state of a person by analyzing his brain signals recorded during listening to the Quran and using a machine learning approach. It is assumed that listening to the Quran brings reverence, and hence two types of emotions emerge which are distinguished as happy and unhappy. In our analysis, we used the Power Spectral Density (PSD) of different bands as features and the Support Vector Machine (SVM) as a classifier. Experiments were conducted by 14 participants and they gave a classification accuracy rate 85.86%.