非线性时间序列ARMA模型的优化估计法

提供了一种ARMA模型参数的优化估计法一阻尼最小二乘法，它结合了Newton法和最速下降法的优点，既保证了迭代计算的收敛性，又加快了收敛的速度。当初值的精度较差时，更宜采用阻尼最小二乘法，而且本文给出实例的MATLAB程序，并利用t统计量检验出：阻尼最小二乘法要比最小二乘法的参数估计值更为显著，拟合模型更优。

An Optimum Algorithm for Nonlinear Time Series ARMA Parameter Estimation

This paper presents an optimum estimation algorithm, damped least square method, combining the
merit of Newton combines with the that of the rapidest descend for nonlinear time series ARMA parameter estimation. The convergence of the iterative computations is ensured and accelerated. This method is suitable if the precision of the initial value is bombing. Matlab for estimating the parameters of an example is given,which shows that the damped least square method is more remarkable than the least square method by t-test.