基于相关向量机的大豆生长发育阶段预测  

Prediction of Soybean Growth and Development Stages Based on Relevance Vector Machine

作　　者：詹环[1] 董玉才[1] 赵海清[2] 鞠桂玲[1] 闵详娟[1] 

机构地区：[1]装甲兵工程学院非线性研究所,北京100072 [2]湛江师范学院数学与计算科学学院,广东湛江524088

出　　处：《嘉应学院学报》2010年第5期27-30,共4页Journal of Jiaying University

摘　　要：精确地预测大豆的生长发育阶段对于农业生产和研究都具有重要的意义,采用相关向量机方法为大豆的生长发育阶段预测建模,通过优化建模参数,所建相关向量机模型具有较高的拟合能力,且预测误差小,稳健性好。实验结果表明,相关向量机建立的模型优于神经网络。Accurate prediction of soybean growth and development stages has very important sense in agricultural industry and research. Relevance Vector Machine （RVM） was applied to set up the model of soybean growth and development stages in this paper. By optimizing the model parameters,the obtained RVM model has high fitting abilities,less prediction errors and less standard deviation of prediction errors. Experimental results indicate that the model set up by RVM out-performed artificial neural networks.

关 键 词：生长期 成熟期组 相关向量机 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]