Nonparametric function estimation for time series by local average estimators (Q688404)

The author considers the nonparametric estimation of the conditional mean function of a stationary time series, where it is assumed that a strong mixing property in the locally transitive sense is satisfied. This dependence assumption is weaker than many other dependence conditions used in time series models. It is shown that under appropriate regularity conditions local average estimators of the conditional mean function achieve the optimal rate of weak convergence in the sup-norm restricted to a compact subset.