Project goals

The use of models with latent variables is important in many diciplines, and also in the management of marine resources. Catch-at-age models used in fisheries stock assessments may contain several hundred unknown parameters, but the are typically only a few parameter of real interest (e.g. current stock size). We will develop method for estimating interest parameters by maximum likelihood, with the remaining unknown parameters modelled as latent variables integrated out of the likelihood function. Numerical integration in high dimensions is challenging, particularly in combination with optimization, and new efficient optimization algorithms in conjunction with high dimensional numerical integration are required. A prototype of a statistical software system for fitting latent variable models will be developed. Estimation is only one part of statistical inference. We will also develop practical diagnostic tools for latent variable models, and we will develop simulation-based methods for estimating bias-corrected confidence distributions. Finally, we will communicate the results from the project to people working in assessment of marine recourses.