农业研究中非线性模型参数估计的贝叶斯方法  

Bayes Method for Parameter Estimation of non-Linear Model in Agriculture Research

作　　者：李小蔚[1] 刘金剑[2] 朱韶洁 王亚民[3] 

机构地区：[1]连云港职业技术学院人事处,江苏连云港222006 [2]连云港职业技术学院商学院,江苏连云港222006 [3]连云港职业技术学院基础部,江苏连云港222006

出　　处：《连云港师范高等专科学校学报》2012年第4期92-94,97,共4页Journal of Lianyungang Normal College

基　　金：江苏省高等学校大学生实践创新训练计划项目(编号:2012JSSPITP4270);江苏省"青蓝工程"中青年学术带头人培养基金(苏教[2010]27号);江苏省第四期"333高层次人才培养工程"资助(苏办发[2011]8号)

摘　　要：在农业研究中,非线性模型有着广泛的应用,其参数估计往往采用变量替换等经典方法。该研究以非线性模型为例,提出了基于Gibbs抽样的贝叶斯估计方法,并以农业研究中的一个实例,演示了贝叶斯估计方法的可行性。结果表明:非线性模型参数估计的贝叶斯方法稳健可靠,适用于农业研究中的复杂非线性模型。In agriculture research, non - linear models were used widely. The method of variable transformation was common used method for the parameter estimation of these non- linear models. In current research, taken as an example, a Bayes method based on the Gibbs sampling approach was proposed for the parameter estimation of the non - linear model. The va- lidity of this method was presented by a real data in agriculture research. The results showed that the proposed Bayes method was effective and suitable for the complex non - linear models in agriculture research.

关 键 词：贝叶斯估计 非线性回归模型 GIBBS抽样 

分 类 号：S126[农业科学—农业基础科学]