论文部分内容阅读
SVM有着良好的分类性能,在各种分类实践中得到了广泛应用。针对SVM的关键参数,如惩罚因子C和核函数参数γ选取不当会影响SVM性能的问题,提出采用收敛速度快、寻优精度高的萤火虫算法(FA)对SVM关键参数进行自动寻优,建立FA-SVM分类模型,并将该模型应用于脉搏信号的情感识别中。情感分类结果表明,SVM经FA算法优化后,对样本分类的识别率比未经任何处理的SVM高出7.9%,验证了该方法的有效性。
SVM has good classification performance, has been widely used in various classification practices. According to the key parameters of SVM, such as improper selection of penalty factor C and kernel function γ, the performance of SVM is affected. The firefly algorithm (FA) with fast convergence rate and high precision is proposed to search the key parameters of SVM automatically. FA-SVM classification model, and the model is applied to the emotional recognition of pulse signals. The results of emotion classification show that the SVM is optimized by the FA algorithm, and the recognition rate of the sample classification is 7.9% higher than that of the SVM without any processing, which verifies the effectiveness of the proposed method.