Mental health is an important part of our physical health and the foundation for us to have a happy life.The high pressure and fast-paced life in modern society make people increasingly prone to anxiety,depression and emotional instability.Reasonable and healthy emotion detection methods are beneficial to help medical staff supervise and regulate the emotions of patients.Based on the existing two wearable methods of electrocardiogram(ECG)and pulse wave(PPG)for monitoring emotions,this paper uses the WESAD dataset to separately verify the importance of heart rate and heart rate specificity of electrocardiogram signals for four emo-tional states:neutrality,stress,pleasure and meditation.Based on the classifier of Sklearn library,the cross-validation method is used to evaluate and compare their performances.The experimental results show that the ExtraTrees classification model performs best in the classification of emotional states,and ECG has a better performance in the classifier.
维普
万方数据