BROWSE
Detail
Conference Paper
Three Differential Emotion Classification by Machine Learning Algorithms Using Physiological Signals
Cited - time in 
Share 
Share
- Authors
- Eun-Hye Jang, Byoung-Jun Park, Sang-Hyeob Kim, Jin-Hun Sohn
- Issue Date
- 2012-02
- Citation
- International Conference on Agents and Artificial Intelligence (ICAART) 2012, pp.1-4
- Language
- English
- Type
- Conference Paper
- Abstract
- In HCI researches, human emotion classification has done by machine learning algorithms based on physiological signals. The aim of this study is to classify three different emotional states (boredom, pain, and surprise) by 5 machine learning algorithms using features extracted from physiological signals. 200 college students participated in this experiment. The audio-visual film clips were used to provoke emotions and were tested their appropriateness and effectiveness. EDA, ECG, PPG, and SKT as physiological signals were acquired for 1 minute before each emotional state as baseline and for 1-1.5 minutes during emotional state and were analyzed for 30 seconds from the baseline and the emotional state. 23 parameters were extracted from these signals: SCL, NSCR, mean SCR, mean SKT, maximum SKT, sum of negative SKT, and sum of positive SKT, mean PPG, mean RR interval, standard deviation RR interval, mean BPM, RMSSD, NN50, percenet of NN50, SD1, SD2, CSI, CVI, LF, HF, nLF, nHF, and LF/HF ratio. For emotion classification, the difference values of each feature subtracting baseline from the emotional state were used for analysis using 5 machine learning algorithms. The result showed that an accuracy of emotion classification by SOM was lowest and SVM was highest. This could help emotion recognition studies lead to better chance to recognize various human emotions by using physiological signals. Also, it is able to be applied on human-computer interaction system for emotion detection.
- KSP Keywords
- Audio-visual, College students, Emotion Recognition, Emotional states, Film clips, Human computer interaction, Human emotions, Interaction System, Machine Learning Algorithms, Physiological signals, RR interval