12 Nov
Perry deValpine
Assistant Professor, Environmental Science, Policy, and Management, College of Natural Resources, University of California Berkeley
http://nature.berkeley.edu/~perrydev/
pdevalpine@berkeley.edu
Estimating Ecological Population Dynamics Models: the Good, the Bad, the Nonparametric
Abstract
State-space models of population dynamics incorporate both environmental stochasticity in population dynamics and sampling variability in data. In many applications, ecologists use a Bayesian framework to estimate state-space models. I will give an overview of practical methods for classical maximum likelihood analysis of state-space models and show how this can be complementary to Bayesian analysis. I will illustrate the pitfalls of complex model-fitting methods by showing how the popular errors-in-variables method can give worse results as you get more data. Finally I will introduce semi-parametric state-space models as a way to let the data "speak for themselves" about the dynamic process that generated them.
Bio
Perry de Valpine received his Ph.D. in Ecology from the University of California, Davis. He did post-doctoral work at the National Center for Ecological Analysis and Synthesis, which is located at UC Santa Barbara, and at UC Berkeley. He now as an Assistant Professor in the Department of Environmental Science, Policy and Management at UC Berkeley.
Aquatic & Fishery Sciences Home
The University of Washington is committed to providing access, equal opportunity and reasonable accommodation in its services, programs, activities, education and employment for individuals with disabilities. To request disability accommodation, contact the Disability Services Office at least 10 days in advance at 206-543-6450/V, 206-543-6452/TTY, 206-685-7264 (FAX); dso@u.washington.edu.
webmaster@fish.washington.edu
Updated