Seminar on Probability and Statistics

Seminar information archive ~03/28Next seminarFuture seminars 03/29~

Organizer(s) Nakahiro Yoshida, Teppei Ogihara, Yuta Koike

2021/03/24

14:30-16:00   Online
Register at least 3 days before at the reference URL. The URL for participation sent before the seminar.
Rachel Fewster (University of Auckland)
Stochastic modelling in ecology: why is it interesting? (ENGLISH)
[ Abstract ]
Asia-Pacific Seminar in Probability and Statistics https://sites.google.com/view/apsps/home

The ecological sciences offer rich pickings for stochastic modellers. There is currently an abundance of new technologies for monitoring wildlife and biodiversity, for which no practicable data-analysis methods exist. Often, modelling approaches that are motivated by a specific problem with relatively narrow focus can turn out to have surprisingly broad application elsewhere. As the generality of the problem structure becomes clear, this can also motivate new statistical theory.

I will describe some ecological modelling scenarios that have led to interesting developments from methodological and theoretical perspectives. As time allows, these will include: saddlepoint approximations for dealing with data corrupted by non-invertible linear transformations; information theory for assuring that it is a good idea to unite data from multiple sources; and methods for dealing with so-called 'enigmatic' data from remote sensors, involving a blend of ideas from point processes, queuing theory, and trigonometry. All scenarios will be generously illustrated with pictures of charismatic wildlife.

This talk covers joint work with numerous collaborators, especially Joey Wei Zhang, Mark Bravington, Peter Jupp, Jesse Goodman, Martin Hazelton, Godrick Oketch, Ben Stevenson, David Borchers, Paul van Dam-Bates, and Stephen Marsland.
[ Reference URL ]
https://docs.google.com/forms/d/e/1FAIpQLSf05P9fCZ5Wkasc7clW1XBpkeONPSjPKuCkNYb3oIqnOAu5Mg/viewform