The algebraic structure of moving average time processes
Author:
O.D. Anderson a
| Affiliation: | a Department of Mathematics, Nottingham University, London, England |
DOI:
10.1080/02331887808801423
Publication Frequency:
6 issues per year
Subjects:
Mathematical Statistics;
Statistical Theory & Methods;
Statistics;
Statistics for the Biological Sciences;
Stochastic Models & Processes;
Formats available:
PDF
(English)
View Article:
View Article (PDF)
AbstractThe class of moving average processes of order q is shown to be both a vector space of dimension q+ 1 and an integral domain, under appropriate compositions; and some associated implications for practical time series analysis are discussed. It is also shown that any particular class of uncorrelated mixed autoregressive-moving average processes of order (p=0 or 1q≧0) is closed under multiplication. |
| Keywords: Autocorrelation; Autocovariance matrices; Decomposition of Processes; Integral Domain; Mixed Processes; Moving Average Processes; Seasonal Models; Time Series Analysis; Transformation of Processes; Vector Space |
| view references (7) |

Download Citation


CiteULike
Del.icio.us
BibSonomy
Connotea