EEG-Based Emotion Recognition by Using Machine Learning and Deep Learning

At present, there are many classification methods for emotion recognition. Based on the SEED dataset, this paper explored the recognition performance of three emotion recognition models: support vector machine(SVM), random forest(RF) and convolutional neural network(CNN). Firstly, we compared the accuracy of the five features of power spectral density (PSD), differential entropy (DE), differential asymmetry (DASM), rational asymmetry (RASM), and the differences between DE features of 23 pairs of frontal-posterior electrodes (DCAU) in total frequency bands. It is proved that the DE feature is more conducive to emotion recognition. Secondly, we also compared the accuracy of DE feature in five different frequency bands (Delta, Theta, Alpha, Beta and Gamma). Here we verified that the performance of the Gamma and Beta frequency bands were better than the others. Then, we traversed the recognition accuracy of each channel through the RF method, and chose four different channel combinations for 4, 6, 9 and 12 channels. Finally, we compared the DE features of these four combined channels with the performance of all 62 channels, and it is demonstrated that the accuracy of the four combined channels were comparatively steady, and the best recognition rate was 89.20%, which was similar to the recognition accuracy of the original 62 channels.