In this study, we discuss the comparative results of emotion classification by several algorithms, which classify three different emotional states (happiness, neutral, and surprise) using physiological features. 300 students participated in this experiment. While three kinds of emotional stimuli are presented to participants, physiological signal responses (EDA, SKT, ECG, RESP, and PPG) were measured. Participants rated their own feelings to emotional stimuli on emotional assessment scale after presentation of emotional stimuli. The emotional stimuli had 96% validity and 5.8 point efficiency on average. There were significant differences of autonomic nervous system responses among three emotions by statistical analysis. The classification of three differential emotions was carried out by Fisher's linear discriminant (FLD), Support Vector Machine (SVM), and Neural Networks (NN) using difference value, which subtracts baseline from emotional state. The result of FLD showed that the accuracy of classification in three different emotions was 77.3%. 72.3% and 42.3% have obtained as the accuracy of classification by SVM and NN, respectively. This study confirmed that the three emotions can be better classified by FLD using various physiological features than SVM and NN. Further study may need to get those results to obtain more stability and reliability, as comparing with the accuracy of emotions classification by using other algorithms.
展开▼