基于支持向量机的浮游植物密度预测研究  

Research on prediction of phytoplankton's density using support vector machines

作　　者：冯剑丰[1] 王洪礼[2] 李胜朋[2] 

机构地区：[1]天津大学环境科学与工程学院,天津300072 [2]天津大学机械工程学院,天津300072

出　　处：《海洋环境科学》2007年第5期438-441,共4页Marine Environmental Science

基　　金：国家自然科学基金资助项目(10472077)

摘　　要：利用支持向量机方法对具有非线性及突发性特点的浮游植物密度进行了预测,同时与人工神经网络方法预测的结果进行了比较。结果表明,无论是拟和能力还是预测能力,支持向量机方法都明显优于人工神经网络方法,支持向量机方法比较适合于具有小样本、非线性特点的浮游植物密度预测研究。The theory of support vector machine was used in the prediction of the phytoplankton＇s density with the characteristics of catastrophic and nonlinearity. Furthermore, the predicted result of support vector machine was compared with the result of artificial neutral network. The results showed that the regressed and predicted result of support vector machine was better than artificial neutral network. The theory of support vector machine was fitted for predicting the phytoplankton＇s density with few data and nonlinear.

关 键 词：支持向量机 浮游植物 预测 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程] Q179[自动化与计算机技术—控制科学与工程]