摘要
针对常用分类算法中支持向量机的分类效率低和分类模型的预测性能受到核参数和惩罚因子的影响这一问题,研究一种改进的差分进化算法,优化支持向量机的核参数和惩罚因子,提高支持向量机的分类性能。实验表明,该算法不仅可以有效避免差分进化算法易早熟收敛的问题,更重要的是它可以充分提高支持向量机的分类准确率。
In view of that low classification efficiency of support vector machine and the influence of the prediction performance of classification mod el in common classification algorithms,an improved algorithm for differential evolution is study,which optimizes the kernel parameters and punishment factors of support vector machine and improves the classification performance of support vector machine.Experiments shows that the proposed algorithm not only effectively can avoid the premature and convergence of differential evolution algorithm,but also effectively improve the classification accuracy of support vector machine.
作者
王珂
周瑶
赵媛媛
WANG Ke;ZHOU Yao;ZHAO Yuan-yuan(School of Information and Control Engineering,Xi'an University of Architecture and Technology,Xi'an 710055)
出处
《现代计算机》
2019年第26期8-12,共5页
Modern Computer
