Random self-decomposability and autoregressive processes
We introduce the notion of random self-decomposability and discuss its relation to the concepts of self-decomposability and geometric infinite divisibility. We present its connection with time series autoregressive schemes with a regression coefficient that randomly turns on and off. In particular, we provide a characterization of random self-decomposability as well as that of marginal distributions of stationary time series that follow this scheme. Our results settle an open question related to the existence of such processes.
| Year of publication: |
2010
|
|---|---|
| Authors: | Kozubowski, Tomasz J. ; Podgórski, Krzysztof |
| Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 80.2010, 21-22, p. 1606-1611
|
| Publisher: |
Elsevier |
| Keywords: | Geometric infinite divisibility Geometric stable law Laplace distribution Linnik distribution Non-Gaussian time series |
Saved in:
Saved in favorites
Similar items by person
-
Matrix gamma distributions and related stochastic processes
Kozubowski, Tomasz J., (2022)
-
A bivariate Lévy process with negative binomial and gamma marginals
Kozubowski, Tomasz J., (2008)
-
Matrix gamma distributions and related stochastic processes
Kozubowski, Tomasz J., (2022)
- More ...