Statistical Inference Theory And Application Of Semi-varying Coefficient Models

The varying-coefficient model is defined by the following model:Where are unknown univariate coefficient,εis random error.For given covariates and response variable Y with In order to simplify the model, the invertigators proposed the following model when they test if some coefficient functions are constants.Semivarying coefficients model is an useful extension of the varying coefficient model.In this paper,we discussed mainly the semivarying coefficients model.In chapter 1,we provide some normal methods in nonparametric regression and dwell on the current researches of varying coefficient models and semivarying coefficient models.In chapter 2, We give the the estimation of coefficient functions of semivarying-coeff icient models by Local Polyomial method ,and give the asympototic normality of the estimation .The paper makes use of semivarying-coeff icient models modeling and forecasting the total value in tertiary industry . Offer areliable suggestion for economists and government to make economical policy and forecast Tertiary in chinaIn chapter 3, semivarying-coeff icient models whose error sequence mixing are disdussed. Givng the the estimation of coefficient functions of semivarying-coeff icient models by Local Polyomial method ,and based on this the consistency for the estimation of the coefficient functions are studied.