On recursive estimation for time varying autoregressive processes

Abstract

This paper focuses on recursive estimation of time varying autoregressive processes in a nonparametric setting. The stability of the model is revisited and uniform results are provided when the time-varying autoregressive parameters belong to appropriate smoothness classes. An adequate normalization for the correction term used in the recursive estimation procedure allows for very mild assumptions on the innovations distributions. The rate of convergence of the pointwise estimates is shown to be minimax in $\beta$-Lipschitz classes for $0<\beta \le 1$. For $1<\beta \le 2$, this property no longer holds. This can be seen by using an asymptotic expansion of the estimation error. A bias reduction method is then proposed for recovering the minimax rate.

Published at http://dx.doi.org/10.1214/009053605000000624 in the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

Cited by in corpus (7)

• Nonparametric regression for locally stationary time series
• Inference of time-varying regression models
• Empirical spectral processes for locally stationary time series
• A recursive online algorithm for the estimation of time-varying ARCH parameters
• Aggregation of predictors for nonstationary sub-linear processes and online adaptive forecasting of time varying autoregressive processes
• Functional estimation and change detection for nonstationary time series
• Joint Covariance Estimation with Mutual Linear Structure