改进的蚁群系统在心电信号情感识别中的应用

Application of improved Ant Colony System for emotion recognition from Electrocardiogram signals

针对心电信号(Electrocardiogram,ECG)的情感识别问题,将局部搜索策略和变异引入蚁群系统(Ant Colony System,ACS)用于特征选择;用K近邻法对高兴和悲伤两种情感分类,在获得较高的识别率和有效特征组合的同时,提高了收敛速度,最好识别率达到93.64%。实验仿真结果表明,该方法是行之有效的。

For the problem of emotion recognition of Electrocardiogram（ECG） signals,a method of feature selection based on Ant Colony System（ACS） that has introduced local search strategy and variation is used to select feature,which accelerates convergence and obtains higher recognition rate and effective feature subset, and the best recognition rate can reach 93.64% using K-nearest neighbor for emotion classification.Simulation results show that this method is feasible.