基于聚类PSO-LSSVM模型的PAD维度预测

Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-28 Cooperative journals: 《计算机应用研究》

胡艳香 孙颖 张雪英 段淑斐

Abstract： In view of the imprecision problem for PAD(Pleasure、Arousal、Dominance) prediction, this paper proposed clustering PSO-LSSVM model combineing Least Squares Support Vector Machine(LSSVM) optimized by Particle Swarm Optimization(PSO) and affective clustering analysis. Firstly, selecting three emotion speeches of TYUT2.0 emotional speech database and Berlin voice library, and extracting emotion features. Establishing Single emotional dimension PSO-LSSVM models for three single emotion and the mixed emotion dimension PSO-LSSVM model for three emotions based on emotion features and P, A and D values. The mothod used mixed emotion dimension PSO-LSSVM model to predict the P, A and D values of the test set, and calculated the distance between the predictive PAD and the PAD of the basic emotion. Finally clustering the emotion whose distance is greater than the threshold into mixed emotion, and clustering the emotion whose distance is less than the threshold into the nearest emotions, then using the corresponding emotional dimension regression model to predict its P, A and D. The research showed that the predictive error of clustering PSO-LSSVM regression model to P, A and D was smaller than that of LSSVM and PSO-LSSVM model, and the correlation between the predicted value and the tagged value was stronger. So the clustering PSO-LSSVM regression model is more reliable and accurate in predicting P, A and D values.

 Hits 1761  Downloads 870  Comment 0