Quantcast

B-spline estimation for varying coefficient regression with spatial data

Research paper by QingGuo Tang, LongSheng Cheng

Indexed on: 04 Dec '09Published on: 04 Dec '09Published in: Science in China. Series A, Mathematics, physics, astronomy / Chinese Academy of Sciences



Abstract

This paper considers a nonparametric varying coefficient regression with spatial data. A global smoothing procedure is developed by using B-spline function approximations for estimating the coefficient functions. Under mild regularity assumptions, the global convergence rates of the B-spline estimators of the unknown coefficient functions are established. Asymptotic results show that our B-spline estimators achieve the optimal convergence rate. The asymptotic distributions of the B-spline estimators of the unknown coefficient functions are also derived. A procedure for selecting smoothing parameters is given. Finite sample properties of our procedures are studied through Monte Carlo simulations. Application of the proposed method is demonstrated by examining voting behaviors across US counties in the 1980 presidential election.