Barnett, V., Kenward, Michael G. (1998) Testing a Poisson renewal process in the context of security alarm maintenance policies. Communications in Statistics-Theory and Methods, 27 (12). pp. 3085-94. ISSN 0361-0926. (The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided) (KAR id:17572)
The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided. |
Abstract
The results of Barnett and Kenward (1996) on alarm and inspection calls for security systems are extended. Some further theoretical properties of the relevant superimposed renewal process model are obtained for the case of a Poisson alarm process both when the inter-inspection interval is constant and when it takes the form of another independent Poisson process. These properties are used to motivate the choice of inference procedures for examining the basic nature of the underlying processes. The original data of Barnett and Kenward (1996) are analysed using such procedures as well as extensive further data. It is seen that the conventional simple assumptions for the inter-inspection intervals do not hold in practice, and consequently a conditional approach to the problem using the Dirichlet distribution is developed and applied.
Item Type: | Article |
---|---|
Uncontrolled keywords: | Dirichlet distribution; heterogenous Poisson process; inspection paradox; maintenance; security and burglar alarms |
Subjects: | Q Science > QA Mathematics (inc Computing science) > QA276 Mathematical statistics |
Divisions: | Divisions > Division of Computing, Engineering and Mathematical Sciences > School of Mathematics, Statistics and Actuarial Science |
Depositing User: | I. Ghose |
Date Deposited: | 05 Apr 2009 07:12 UTC |
Last Modified: | 05 Nov 2024 09:53 UTC |
Resource URI: | https://kar.kent.ac.uk/id/eprint/17572 (The current URI for this page, for reference purposes) |
- Export to:
- RefWorks
- EPrints3 XML
- BibTeX
- CSV
- Depositors only (login required):