Bootstrapping the Kalman Filter.
Abstract
The bootstrap is proposed as a method for estimating the precision of forecasts and estimates of parameters of the Kalman Filter model. It is shown that when the system and the filter is in steady state the bootstrap applied to the Gaussian innovations yields asymptotically consistent standard errors. That the bootstrap works well with moderate sample sizes and supplies robustness against departures from normality is substantiated by empirical evidence. Keywords: Bootstrap; Kalman filter; Forecasting; and Robustness.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1984
- Accession Number
- ADA150509
Entities
People
- D. S. Stoffer
Organizations
- University of Pittsburgh