摘要
建立了一种支持向量机(SVM)的水资源质量分类评价模型,采用"一对一"的多类别分类算法,核函数取径向基函数,分别用网格搜索法、遗传算法和粒子群算法对SVM模型的控制参数进行寻优.农村水质分类评价实验表明:网格搜索方法能得到较高的分类准确率而且泛化能力较强,计算时间短等优点,该模型的实际应用可以推广.
A water quality classification and evaluation model based on support vector machine （SVM） was es- tablished. Multi class classification algorithm was designed by using ＂one against one＂ method and taking radial basis function （RBF） as kernel function. Grid search method, genetic algorithm （GA） and particle swarm optimization （PSO） algorithm were used to optimize the control parameters of SVM. The rural water quality classification evaluation experiment show that Grid search method can achieve higher classification accuracy and better generalization ability, and the computation time is shortest. The practical application of the SVM model can be popularized.
出处
《邵阳学院学报（自然科学版）》
2016年第2期11-15,共5页
Journal of Shaoyang University：Natural Science Edition
基金
邵阳市科技计划项目(2015NC43)
关键词
支持向量机
水质分类评价
网格搜索
参数寻优
support vector machine （SVM）
classification and evaluation of water quality
grid search
parameter optimization
