Presentation
Nonparametric Transfer Function Model
ICSA Applied Statistics Symposium, University of Connecticut
(2006)
Abstract
A new approach is proposed to model the relationship between an output and some input time series, perturbed by correlated noise. The functional form of this relationship (the transfer function) is unknown but assumed to be smooth. We propose to model the transfer function by nonparametric smoothing methods and model the noise as an ARIMA process. By using nonparametric smoothing, the model is very flexible and can be used to model highly nonlinear relationships of unknown form; by modeling the noise, the correlation in the data is removed so the transfer function can be estimated more efficiently; additionally, the estimated ARIMA structure can be used to improve the forecasting performance. The estimation procedures are introduced and the asymptotic properties of the estimators are studied. The finite-sample properties of the estimators are studied by simulation and real-life examples.
Keywords
- Nonparametric regression,
- Local polynomial regression,
- Regression splines,
- Nonlinear time series analysis,
- Transfer function model,
- ARIMA
Disciplines
Publication Date
2006
Citation Information
Jun Liu. "Nonparametric Transfer Function Model" ICSA Applied Statistics Symposium, University of Connecticut (2006) Available at: http://works.bepress.com/jun_liu/27/