A0506
Title: Computational efficiency and precision for replicated-count and batch-marked hidden population models
Authors: Matthew Parker - Simon Fraser University (Canada) [presenting]
Laura Cowen - ()
Jiguo Cao - Simon Fraser University (Canada)
Lloyd Elliott - Simon Fraser University (Canada)
Abstract: Two computational issues are addressed, common to open-population N-mixture models, hidden integer-valued autoregressive models, and some hidden Markov models. The first issue is computation time, which can be dramatically improved through the use of a fast Fourier transform. The second issue is the tractability of the model likelihood function for large numbers of hidden states, which can be solved by improving the numerical stability of calculations. As an illustrative example, the application of these methods is detailed in the open-population N-mixture models. Computational efficiency and precision are compared between these methods and standard methods employed by state-of-the-art ecological software. Faster computing times are shown (a ~6 to ~30 times speed improvement for population size upper bounds of 500 and 1000, respectively) over state-of-the-art ecological software for N-mixture models. The methods are applied to compute the size of a large elk population using an N-mixture model, and it is shown that while the methods converge, previous software cannot produce estimates due to numerical issues. These solutions can be applied to many ecological models to improve precision when logs of sums exist in the likelihood function and to improve computational efficiency when convolutions are present in the likelihood function.
