将遗传算法应用到生理信号情绪识别的特征选择研究过程中。对与情绪相关的四种生理信号进行特征选择,并用支持向量机进行分类评估,得出基于最优识别率的特征子集,为后续进行基于生理信号的情绪状态分类器的设计与研究奠定基础。 The genetic algorithm is applied to the process of feature selection of physiological signal emotion recognition. The four physiological signals related to emotions were selected by feature selection, and then the SVM was used to classify the features to get the feature subset based on the optimal recognition rate, which laid the foundation for the subsequent design and research of emotional state classifier based on physiological signals .