基于人工蜂群算法的支持向量机参数优化及应用

为了解决常用的支持向量机(SVM)参数优化方法在寻优过程不同程度的陷入局部最优解的问题,提出一种基于人工蜂群(ABC)算法的SVM参数优化方法。将SVM的惩罚因子和核函数参数作为食物源位置,分类正确率作为适应度,利用ABC算法寻找适应度最高的食物源位置。利用4个标准数据集,将其与遗传(GA)算法、蚁群(ACO)算法、标准粒子群(PSO)算法优化的SVM进行性能比较,结果表明,本文方法能克服局部最优解,获得更高的分类正确率,并在小数目分类问题上有效降低运行时间。将本文方法运用到计算机笔迹鉴别,对提取的笔迹特征进行分类,与GA算法、ACO算法、PSO算法优化的SVM相比,得到了更高的分类正确率。

SVM parameter optimization and application based on artificial bee colony algorithm

In order to solve the difficult problem of falling into local optimal solution which all the common support vector machine(SVM) parameter optimization methods have in different degree,a new SVM parameter optimization method based on artificial bee colony(ABC) algorithm is proposed and applied to computer handwriting verification.Penalty factor C and kernel function parameter σ of SVM were taken as the optimization objects,and the classification accuracy of SVM was used as the fittness value.Tests on four standard datasets show that compared with colony algorithm,practical swarm algorithm and genetic algorithm,the proposed method overcomes the local optimal solution problem,achieves higher classification accuracy,and decreased the running time efficiently in small number classification problems.Then the proposed method was applied to handwriting verification.Compared with the SVM optimized by the other three optimization algorithms,the method reposed in this paper obtains higher classification accuracy.