Computing observation weights for signal extraction and & ltering

  title={Computing observation weights for signal extraction and & ltering},
  author={Siem Jan Koopmana and Andrew Harveyb},
  • Siem Jan Koopmana, Andrew Harveyb
  • Published 1999
In this note we present an algorithm for computing the weights assigned to observations when carrying out prediction, ltering and signal extraction using a model in state space form. In linear time-invariant models such weights can be obtained analytically from the WienerKolmogorov formulae. However, our method is general, being applicable to models with deterministic components, time-varying state matrices and multivariate models. Weight patterns can be compared with kernels typically used in… CONTINUE READING
Highly Cited
This paper has 38 citations. REVIEW CITATIONS


Publications citing this paper.
Showing 1-10 of 28 extracted citations


Publications referenced by this paper.
Showing 1-10 of 17 references

Messy Time Series

T. B. Fomby, R. C. Hill
Advances in Econometrics . Volume 13 • 1998

Weights and Robustness of Model - based Seasonal Adjustment

M. Riani

Exact initial Kalman ltering and smoothing for nonstationary time series models

S. J. Koopman
Journal of the American Statistical Association • 1997

The simulation smoother for time series models

P. de Jong, N. Shephard
Biometrika • 1995

Detecting level shifts in time series

N. S. Balke
Journal of Business and Economic Statistics • 1993

The Di use Kalman lter

P. de Jong
Annals of Statistics • 1991

Filtering and smoothing in state space models with partially di use initial conditions

C. F. Ansley, R. Kohn
Journal of Time Series Analysis • 1990

A fast algorithm for signal extraction , inFuence and cross - validation

S. J. Koopman, N. Shephard, J. A. Doornik
Biometrika • 1989

Similar Papers

Loading similar papers…