Publication: Un método de inicialización del filtrado para modelos en espacio de los estados con inputs estocásticos
Loading...
Official URL
Full text at PDC
Publication Date
1996-09
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Facultad de Ciencias Económicas y Empresariales. Instituto Complutense de Análisis Económico (ICAE)
Citations
Exportar
Abstract
En este trabajo se derivan las expresiones exactas de la media y varianza condicional del estado inicial de un modelo en espacio de los estados con inputs estocásticos, generalizando los resultados teóricos obtenidos por De Jong y Chu-Chun-Lin (1994). Se muestra que las condiciones iniciales exactas dependen del carácter estacionario o no estacionario del modelo y que las estimaciones finales de los parámetros son sensibles a la presencia de inputs estocásticos, siendo ésta una situación frecuente en Econometría.
We derive exact expressions for the conditional mean and variance of the initial state of a state space system with stochastic inputs, under stationarity or nonstationarity. These results generalize those of De Jong and Chu-Chun-Lin (1994) and provide a useful initialization method to obtain maximum likelihood estimates of the model parameters. As final estimates are sensitive to initial conditions, the presence of stochastic inputs -a frequent situation ín Econometrics- should be considered when computing the mean and variance of the initial state.
We derive exact expressions for the conditional mean and variance of the initial state of a state space system with stochastic inputs, under stationarity or nonstationarity. These results generalize those of De Jong and Chu-Chun-Lin (1994) and provide a useful initialization method to obtain maximum likelihood estimates of the model parameters. As final estimates are sensitive to initial conditions, the presence of stochastic inputs -a frequent situation ín Econometrics- should be considered when computing the mean and variance of the initial state.
Description
UCM subjects
Unesco subjects
Keywords
Citation
Anderson, B.D.O. yMoore, J.B. (1979). Optimal Filtering. Englewood Cliffs. NJ: Prentice Hall.
Burridge, P. Y Wallis, K.F: (1985).Calculating the variance of seasonally adjusted series. Journal of the American Statistical Association, 80, 541-552.
Danyang, L. y Xuanhuang, L. (1994). Optimal State Estimation without the Requirement of a Priori Statistics Information of the Initial state. IEEE Transactions on Automatic Control, 39, 10, 2087-2091.
De Jong, P. (1988). The likelihood far a State Space Model. Biometrika 15, 1, 165-169.
De Jong, P. Y Chu-Chun-Lin, S. (1994). Stationary and Non-Stationary State Space Models. Journal of Time Series Analysis, 15, 2, 151-166.
García-Ferrer, A., del Hoyo, J., Novales, A. y Young, P.C. (1996). Recursive identification, estimation and forecasting of Nonstationary Economic Time Series with Applications to GNP International Data, en Bayesian Analysis in Statistics and Econometrics, Essays in Honor of Arnold Zellner , (eds., D.A. Berry, K.M. Chaloner and J.K. Geweke), John Wiley and Sons, Inc.
Gardner, G., Harvey, A.C., y Phillips, G.D.A. (1980). Algorithm 154. An Algorithm for Exact Maximum-Likelihood Estimation of Autoregressive-Moving Average Models by Means of Kalman Filtering. Applied Statistics, 29, 311-317.
Kitagawa, G. (1981). A Nonstationary Time Series Model and its Fitting by Recursive Filter. Journal of Time Series Analysis, 2, 103-116.
Marshall, P. (1992). State Space Models with Diffuse Initial Conditions. Journal of Time Series Analysis, 13, 5, 411-414.
Mauricio, J.A. (1995). Exact Maximum Likelihodd Estimation of Stationary Vector ARMA Models. Journal of the American Statistical Association, 90, 429, 282-291.
Petkov, P.Hr., Christov, N.D. y Konstantinov, M.M. (1991). Computational Methods for Linear Control Systems. Prentice-Hall, Englewood Cliffs, N.J.
Rosenberg, B. (1973). The Analysis of a Cross Section of Time Series by Stochastically Convergent Parameter Regression. Annals of Economic and Social Measurement, 2, 4, 399-428.
Shea, B.L. (1987). Estimation of Multivariate Time Series. Journal of Time Series Analys1s, 8, 95-109.
Sotoca, S. (1994). Aplicación del Filtro de Chandrasekhar a la Estimación por Máxima Verosimilitud Exacta de Modelos Dinámicos. Estadística Española, 36, 136, 259 285.
Terceiro, J. (1990). Estimation of Dynamic Econometric Models with Errors in Variables. Springer-Verlag, Heidelberg.