一类函数加性误差模型参数估计的改进

The improvement of parameter estimation about additive error model of a kind of function

主要就函数y=axb加性误差模型对数线性化后的随机误差不满足零均值性和同方差性,提出了一种有效的两阶段加权最小二乘估计方法(TSWLS),并用模拟数据试验检验了理论的正确性.

In view of Zero Mean and Homoskedasticity of the Error of the Linearization Model about additive error model of the Function y=ax^b, the improved estimate by Two Stage WLS method is presented in this paper. It is proved in the imitate experiment that the improved estimate is better than the ordinary least square estimate.