|
Organizers |
Estimation of Population Size in a Poisson Mixture Model
by
Bruce G. Lindsay
Penn State University, USA
Coauthors: Chang Xuan Mao (University of California at Berkeley)
The "number of species" problem has long been a source of statistical interest and research.
We phrase the problem here in terms of a Poisson mixture model where the latent (mixing) distribution is modeled nonparametrically. One can think of the latent variables being the unseen true abundance of the individual species, from which individuals are sampled in a Poisson fashion.
In this setting we can recover the standard methods used in this problem by using moment estimators of population quantities. The model-based framework enables one to extend these methods in a natural way to create for estimating functions that allow arbitrary heterogeneity.
Author's technical report site
Date received: November 2, 2001
Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caid-57.