Work

The periodically-observed time-homogeneous Poisson process, with applications in violence and loss

Year: 2025
Type: article
Abstract: When analytical tools can show that arrival data for practical systems are indistinguishable from those of a time-homogeneous Poisson process (THPP), important insights can be drawn from these systems; this is particularly true for phenomena of violence and loss, where events occur with minimal warning or apparent rationale.Practical constraints (e.g., manpower, money, technology) frequently prevent the exact arrival data of a system from being collected, and subsequent analysis must be performed on the available approximations of the arrival data.Although a common approach to modeling in the presence of periodic observation is simply to ignore the error associated with these approximations of arrival data, researchers have noted that this impedes proper analysis of these data.Therefore, a practitioner's ability to e!ectively identify THPPs in periodically observed practical systems-and therefore draw important insights about these systems-relies on the availability of analytical tools that can properly accommodate periodically observed data.This research introduces the periodically-observed time-homogeneous Poisson process (PTPP) framework, which explicitly accounts for the presence of periodic observation within THPP systems.This framework enables the study of how the act of periodic observation during data collection can influence insights drawn from THPP systems.First, periodic observation is shown to pose challenges to traditional goodness-of-fit techniques for determining whether arrival data originate from a THPP.This research demonstrates that the implementation of the PTPP model for goodness-of-fit testing can overcome these challenges.Second, theArrival rate function Matrices P PTPP transition matrix Q Order-only transition matrix ix
Language: English
Open Access status: green
OpenAlex ID: W7220757407
FWCI:
Cites: 54
Cited by:
Related to: 10