摘要
提出了基于支持向量机的水环境质量综合评价方法。给出了数据规范化处理及学习样本生成的具体方法,采用“一对一”(1-v-1)的多类别分类算法,使用k-fold交叉验证优化参数。对湖泊水环境富营养化程度的综合评价结果表明,该方法简便、客观,具有较强的实用性。
A method of comprehensive assessment of water environmental quality based on support vector machine is proposed in this paper. The corresponding ways of data standardization and generating training data are given, Multi-class support vector machine was designed by using one-versus-one method, and to get the optimization parameter by using k-fold cross-validation. The assessment results of five main lakes eutrophication showed that this method possesses convenience objectivity and high practicability.
出处
《中国环境监测》
CAS
CSCD
北大核心
2007年第3期81-84,共4页
Environmental Monitoring in China
基金
国家自然科学基金项目(N40101005)
关键词
支持向量机
一对一
水环境质量
综合评价
Support vector machine
One-versus-one
Water environmental quality
Comprehensive assessment
