Several Cases The Estimates And The Nature Of The Semi-parametric Model

Semi-parameter model is established on the basis of a class model of the parameter models and the non- parameter models,It is better than non-parameter models and the parameter models.This paper is most introduced this semi-parameter models of longitudinal data and the estimation of the parameters and the non - parameters.It resolves the problems of estimation about a class of Semi-parameter model of longitudinal data.This innovation:First,For a class of semi - parametric models of longitudinal data,It presented the T estimation.The second,It discussed the pollution datas's parameter estimations and the strong consistenteny.The third,For a class with the smoothness of a semi-parametric model,under the linear restrictive conditions,It discussed the problems of the parameter estimation and hypothesis testing,but also it put forward several proposals pending future research.In order to address the estimated of half-parameter model under pollution datas,a general review of the research results about this model at home and abroad at present,which sum up seven commonly used the estimated methods,which two are based with the self-variables had errors.which one is pue forth on the bases of the errors forms an AR process.As for being a group of data pollution,This semi - parametric model of longitudinal data used the two-step method,The parameters are estimated, and the strong consistency was proved.For a smooth semi -parametric model with a smooth items,It based on the symmetric and asymmetric of smoothess,under the linear conditions of parameter estimation,presented a hypothesis testing,simulated this situation,and analyed the results.