Spatiotemporal dependency structures and network distances in point processes.
Abstract
Earthquakes cause aftershocks; a terrorist attack induces copycats. The key notion here is self-excitement. This phenomenon occurs in wildly different settings and causes events or objects in space and time to cluster: galaxies, molecules, crimes, and even animals. Yet, clusters also form as a result of other driving forces. The PI is interested in how and why these clusters occur; some events occur at a constant background rate and others arise simply because another event took place in close proximity. How proximity is defined in this context is crucial: the path of least resistance between events may not simply be the straight-line distance between them. It is imperative that we identify the underlying mechanisms driving the occurrence and potential effect of events so that we can accurately predict them. Emergency and contingency planning are reliant on knowing what the risk of an event is and why it may have occurred lives are saved by knowing when there is a greater risk of an earthquake or by being able to predict a terrorism event. This project will develop new statistical methodology to identify self- excitement mechanisms in cluster processes and their impact on outbreak intensity, for applications ranging from disaster resourcing to species distribution.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 21, 2022
- Source ID
- FA23862114028XX0
Entities
People
- Charlotte M Jones-Todd
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Auckland