摘要
针对语音特征参数对某类情感具有不确定性的问题,提出一种基于典型相关性分析的改进模糊支持向量机算法,应用于语音情感识别。采用典型相关性分析方法对特征向量进行降维,得到样本的约简向量集,在此约简向量集上建立模糊支持向量机模型判定情感类型。仿真实验结果表明,该方法相比于传统支持向量机法和模糊支持向量机法具有较高的识别准确率。
An improved fuzzy support vector machine algorithm is proposed in this paper in order to solve the non-determinacy of speech feature parameter.Firstly,canonical correlation analysis is utilized to reduce the dimension of feature vectors.And then,fuzzy support machine is trained on the reduced set to make final decision.The experiment results show that,our method has superior classification performance compared with SVM and FSVM.
出处
《重庆科技学院学报（自然科学版）》
CAS
2012年第5期140-142,共3页
Journal of Chongqing University of Science and Technology：Natural Sciences Edition
基金
甘肃省教育厅基金项目(1113-01)
关键词
典型相关性分析
模糊支持向量机
语音情感识别
支持向量机
canonical correlation analysis
fuzzy support vector machine
speech emotion recognition
support vector machine
