基于支持向量机的绿潮灾害影响因素的权重分析.PDFVIP

中国环境科学 2015,35(11) ：3431~3436 China Environmental Science 基于支持向量机的绿潮灾害影响因素的权重分析 何世钧 1*,唐莹莉 1,张 婷 1,李 煜 1,谢圣东 1,于克锋 2,何培民 2,3 (1.上海海洋大学信息学院,上海 201306 ； 2.上海海洋大学水产与生命学院,上海 201306 ；3.上海海洋大学海洋科学研究院,海洋生态环境与修复研究所,上海 201306) 摘要：根据 2012~2013 年南黄海海域绿潮浒苔遥感监测分布面积数据及温度、天气状况、风向、风力、浪高 5 个影响绿潮浒苔扩散的气 候因子,建立了相应的支持向量机回归模型.通过模型中各影响因素权重的变化分析绿潮灾害的发展过程,并与传统的单因素分析法进行 对比,支持向量机回归更能准确得出各影响因素的权重及权重的变化规律.通过对权重变化规律的分析,给出在绿潮发生过程中漂浮、爆 发和消亡阶段的划分依据. 关键词：支持向量机回归；影响因子；权重分析；灾害过程 中图分类号：X55 文献标识码：A 文章编号：1000-6923(2015)11-3431-06 1* 1 Weight analysis of each influence factor of the green tide disaster based on SVM. HE Shi-jun , TANG Ying-li , ZHANG Ting1 1 1 2 2,3 , LI Yu , XIE Sheng-dong , YU Ke-feng , HE Pei-min (1.College of information, Shanghai Ocean University, Shanghai 201306, China ；2.College of Fisheries and Life Sciences, Shanghai Ocean University, Shanghai 201306, China ；3.Institutes of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China). China Environmental Science, 2015,35(11)：3431~3436 Abstract ：According to the green tide algae - Ulvaprolifera remote sensing data of the Southern Yellow Sea in 2012~2013, climate factors including temperature, weather, wind direction, wind force and wave height which effect the Ulvaprolifera diffusion, the corresponding support vector regression model (SVR) was established. By analyzing the development of the disaster using the model’s weight change, the result of SVM was more accurately draw the right weight of each factor and weight conversion lawcompared with traditional single factor analysis method. Then through the analysis for the change rule of weight, give the