Semi-parametric Regression Model To Estimate The Nature Of A Class Of Mixed Sequence

As the most important branch in statistic, regression analysis method has become necessary tool for statistic inference in other branches, but also promoted the research of econometrics. In modern econometrics,there has been an upsurge of interests and effort in nonparametric and semi-parametric econometrics. In this paper, some further properties in semi-parametric are studied on the base of present research.1. Two kinds of regression model are concerned,classical parametric re-gression and nonparametric regression model, making simple introduction for their concept and making comparison for two kinds of regression model, the semi-parametric regression is introduced.2. The regression analysis about financial time series has become the most important in financial econometrics, financial time series often appeared to be mixing sample sequences, several mixing random sequences are concerned and some limit properties such as three-series theorem, almost sure convergence and strong law of large number are obtained.3. Semi-parametric regression model is concerned when error sequences underρmixing sequences, the strong consistency and r-th mean consistency for the estimators ofβand g(·) is investigated, some results on this kind of question are extended.