# Article

Full entry | PDF   (1.1 MB)
Keywords:
improvement of prediction; discrete stationary process
Summary:
Let $\{W_t\}=\{(X'_{t'}, Y'_t)'\}$ be vector ARMA $(m,n)$ processes. Denote by $\hat{X}_t(a)$ the predictor of $X_t$ based on $X_{t-a}, X_{t-a-1}, \ldots$ and by $\hat{X}_t(a,b)$ the predictor of $X_t$ based on $X_{t-a}, X_{t-a-1}, \ldots, Y_{t-b},Y_{t-b-1}, \ldots$. The accuracy of the predictors is measured by $\Delta_X(a)=\text{E}[X_t-\hat{X}_t(a)][X_t-\hat{X}_t(a)]'$ and $\Delta_X(a,b)=\text{E}[X_t-\hat{X}_t(a,b)][X_t-\hat{X}_t(a,b)]'$. A general sufficient condition for the equality $\Delta_X(a)=\Delta_X(a,a)]$ is given in the paper and it is shown that the equality $\Delta_X(1)=\Delta_X(1,1)]$ implies $\Delta_X(a)=\Delta_X(a,a)]$ for all natural numbers $a$.
References:
[1] J. Anděl: Measures of dependence in discrete stationary processes. Math. Operationsforsch. Statist., Ser. Statistics 10 (1979), 107-126. MR 0542367
[2] J. Anděl: On extrapolation in two-dimensional stationary processes. To appear in Math. Operationsforsch. Statist., Ser. Statistics. MR 0596515
[3] J. Anděl: Some measures of dependence in discrete stationary processes. Doctoral dissertation (Department of Statistics, Charles University, Prague, 1980) (in Czech). MR 0542367
[4] T. Cipra: Correlation and improvement of prediction in multivariate stationary processes. Ph. D. dissertation (Department of Statistics, Charles University, Prague, 1980) (in Czech).
[5] T. Cipra: Improvement of prediction in multivariate stationary processes. Kybernetika 17 (1981), 234-243. MR 0628211
[6] T. Cipra: On improvement of prediction in ARMA processes. Math. Operationsforsch. Statist., Ser. Statistics 12(1981), 567-580. MR 0639253 | Zbl 0514.62103
[7] W. A. Fuller: Introduction to statistical time series. Wiley, New York, 1976. MR 0448509 | Zbl 0353.62050
[8] C. W. J. Granger: Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37 (1969), 424-438. DOI 10.2307/1912791
[9] D. A. Pierce L. D. Haugh: Causality in temporal systems. Journal of Econometrics 5 (1977), 265-293. DOI 10.1016/0304-4076(77)90039-2 | MR 0443261
[10] Yu. V. Rozanov: Stationary random processes. Gos. izd., Moskva, 1963 (in Russian).
[11] C. A. Sims: Money, income and causality. American Economic Review 62 (1972), 540-552.

Partner of