Quantile-Wavelet Nonparametric Estimates for Time-Varying Coefficient Models

The paper considers quantile-wavelet estimation for time-varying coefficients by embedding a wavelet kernel into quantile regression. Our methodology is quite general in the sense that we do not require unknown to be smooth curves of common degree or errors independently distributed. Quantile-wavelet robust outliers heavy-tailed data. model dynamic nonlinear time series. A strong Bahadur order O2mn3/4(logn)1/2 obtained under mild conditions. As applications, rate uniform convergence and asymptotic normality are derived.