摘要
在浮游植物总量与环境因子的定量关系研究中,使用了多元自适应样条回归模型。基于2003年5-9月渤海湾地区浮游植物总量及各种环境因子的实测数据,经过与投影寻踪回归模型预报结果对比,表明多元自适应样条回归很好地反映了浮游植物总量与环境因子定量关系并且是预报赤潮的较好模型。
The model of muhivariate adaptive regression splines is suggested in analysis of quantitative association between total number of phytoplankton and environmental actors. Based on the data of harmful algal and environmental actors in Bohai bay, it is shown that multivariate adaptive regression splines is a good method to post quantitative association between total number of phytoplankton and environmental factors and forecast phytoplankton in contrast to projection pursuit regression.
出处
《海洋技术》
2006年第3期7-9,共3页
Ocean Technology
基金
国家自然科学基金资助项目(10472077)
关键词
赤潮
浮游植物
多元自适应回归
投影寻踪回归
Harmful algal
Phytoplankton
Multivariate adaptive regression splines
Projection pursuit regression