Robinson, Peter (1982) Analysis of time series from mixed distributions. Annals of Statistics, 10 (3). pp. 915-925. ISSN 0090-5364
Full text not available from this repository.
Identification Number: 10.1214/aos/1176345881
Abstract
Some stationary and non-stationary time series arise from mixed distributions, the probabilities attached to the occurrence of certain values being positive, while a continuum of possible values is also involved. Such series are modeled in terms of a stationary Gaussian process $X_t$, which is censored when it crosses certain thresholds. Procedures are proposed for estimating the autocorrelation function of $X_t$. Their strong consistency and asymptotic normality are established. We suggest tests of the hypothesis that $X_t$ is white noise.
| Item Type: | Article |
|---|---|
| Official URL: | http://imstat.org/en/index.html |
| Divisions: | LSE |
| Date Deposited: | 27 Apr 2007 |
| Last Modified: | 15 Sep 2023 12:04 |
| URI: | http://eprints.lse.ac.uk/id/eprint/1228 |
Actions (login required)
 |
View Item |
