摘要
针对生理信号的情感识别问题,本文提出了禁忌搜索(TS)和神经网络结合的情感状态分类方法。该方法用禁忌搜索算法选择情感生理信号特征,神经网络BP算法分类情感状态。仿真实验结果表明,BP分类器在TS选择的特征基础上达到很好的情感分类效果。
This paper propose a method of TS-MLP about emotion recognition of physiological signal . It can recognize emotion successfully by Tabu search which selects features of emotion＇ s physiological signals and Muhilayer Perceptron that is used to classify emotion. Simulation shows that it has achieved good emotion classification performance.
出处
《齐齐哈尔大学学报（自然科学版）》
2013年第3期19-22,共4页
Journal of Qiqihar University(Natural Science Edition)
关键词
情感识别
禁忌搜索
生理信号
特征选择
多层感知器
emotion recognition
tabu search
physiological signals
feature selection
muhilayer perceptron
