非线性自回归模型误差密度估计的Berry-Esseen界

Berry-Esseen Bound for Error Density Estimatorsin Nonlinear Autoregressive Models

考虑非线性自回归模型X i=rθ(X i-1,…,X i-s)+εi,其中:θ为q维未知参数;{εi}为独立同分布的随机误差,且均值为0、方差为σ2.在适当的假设条件下,给出非线性自回归模型误差密度估计的Berry-Esseen界.

We consider the nonlinear autoregressive model X i=rθ(X i-1,…,X i-s)+εi,whereθis a q-dimensional unknown parameter,{εi}is an independent a nd identically distributed random error with a mean of zero,and a variance ofσ2.We give the Berry-Esseen bound for error density estimator s in nonlinear autoregressive models under appropriate assumptions.