Mixed autoregressive-moving average multivariate processes with time-dependent coefficients
Conditions for mixed autoregressive-moving average processes with time-dependent coefficients to be purely nondeterministic and invertible can be obtained from classical difference equations theory. These conditions involve one-sided Green's functions or their matricial equivalents. A recursive computation of these functions is proposed, which allows one to drop the assumption of nondegeneracy classicaly made about the highest order matrix of difference operators; it constitutes thus a generalized definition of these functions.
Year of publication: |
1978
|
---|---|
Authors: | Hallin, Marc |
Published in: |
Journal of Multivariate Analysis. - Elsevier, ISSN 0047-259X. - Vol. 8.1978, 4, p. 567-572
|
Publisher: |
Elsevier |
Keywords: | Time series difference equations stochastic equations |
Saved in:
Saved in favorites
Similar items by person
-
Nonstationary first-order moving moving average processes : the model-building problem
Hallin, Marc, (1981)
-
Hallin, Marc, (2022)
-
On distribution and quantile functions, ranks and signs in Rd
Hallin, Marc, (2017)
- More ...