Search

This paper presents an approach for automated estimation of human emotions from electroencephalogram data. The used features are principally the Hjorth parameters calculated for theta, alpha, beta and gamma bands taken from certain channels. The classification stage is support vector machine. Since the human emotions are modelled as combinations from physiological elements such as arousal, valence, dominance, liking, etc., these quantities are the classifier's outputs. The best achieved correct classification performance is about 80%.

This paper presents an approach for automated staging of the human sleep. It is based on analysis of two channel electroencephalogram and an electrooculogram. The classifier is trained with two different groups of features separately and in combination as well. Statistic measures of first and higher order serve as features from the first set. The rules of Rechtschaffen and Kales are exploited for extraction of the second group of features. The contextual classifier for the sleep staging combines Support Vector Machine with Hidden Markov Model. This approach is verified and evaluated with an expert annotated database of biomedical signals and the overall accuracy is over 90%.