非线性模型中无信息方差和协方差分量Bayes估计

Bayesian estimation of non-information variance and covariance components in the nonlinear model

采用Bayes方法从无先验信息出发,得到了非线性模型中方差和协方差分量的估计(包含相关系数),最后通过实例解算,结果表明:非线性模型中方差和协方差分量的估计,与ρ的理论值-0.5偏差不大,当没有先验信息时,该方法是可行的。

The estimation of variance and covariance components in the nonlinear model was obtained with the Bayesian approach without prior information. The variance and covarianee components included correlation coefficient. The present Bayesian estimation of variance and covarianee components extended other available estimations in the linear model. Samples of estimation calculations were presented.