一种改进的支持向量机参数优化方法

Improved Optimization Method for Support Vector Machine Parameter

为了保证支持向量机在提高核参数寻优效率的同时,拥有较高的学习精度,深入研究了核参数对支持向量机分类的影响,分析了网格搜索法和双线性搜索法的优缺点,并以此为基础提出了一种改进的参数优化方法。实验结果表明,该算法在保证支持向量机获得较高学习精度的同时能大大缩短参数寻优的时间,证明了该算法的优越性。

In order to improve the optimization efficiency of support vector machine(SVM) parameter, moreover ensuring high learning accuracy, a further research on the influence of kernel parameter for SVM classification was conducted. Based on the analysis of the pros and cons of grid search and bilinear search methods, an improved parameter optimization method was proposed. Experimental results show that the proposed algorithm not only ensures high learning accuracy but also greatly improves parameter optimization efficiency of SVM. The superiority of the proposed algorithm was presented.