Suntory and Toyota International Centres for Economics and Related Disciplines (STICERD) LSE RSS Contact Us YouTube Twitter


Econometrics Paper
Exact Score for Time Series Models in State Space Form (Now published in Biometrika (1992), 79, 4, pp.283-6.)
Siem Jan Koopman and N.G. Shephard  1992
Paper No' EM/1992/241:

Tags: smoothing; kahman filter; em algorithm; unobserved components model; profile likelihood.

The score vector for a time series model which fits into the Gaussian state space form can be approximated by numerically differentiating the log-likelihood. If the parameter vector is of length p, this involves the running of p + 1 Kalman filters. This paper shows the score vector can be computed in a single pass of the Kalman filter and a smoother. For many classes of models this dramatically increases the speed and reliability of algorithms for the numerical maximisation of likelihood.