Ed Boone
Zero-Inflated models for determining fish abundance
Abstract
Determining which variables influence the abundance of a species of fish
is a common problem in Ecology. However, modeling this is often
problematic due to an over abundance of zeros. This talk proposes using
a Zero-Inflated model to model both presence/absence and abundance
simultaneously. In this model we also incorporate variables at
different scales and spatial correlation among sampling sites. This
model is illustrated on an Ecological dataset concerning the Mary and
Albert Rivers in Australia.
Back to Seminar Page
|