Olav did his PhD in statistics during 2012-2016 at University of Oslo under the supervision of associate professor Ida Scheel (University of Oslo) as well as professors Geir Storvik, Bent Natvig, and Arne Huseby. During his PhD studies, he worked on developing models for bycatch, which was then implemented as a user tool for the Surveillance Service North for cost-effective regulation of shrimp fishing in the Barents Sea. Starting from 2016, he has been working as a research scientist and a senior research scientist at the Norwegian Computing Center, where he continues working with developing models for spatio-temporal statistics and applications in fisheries.
Olav Breivik: Template Model Builder and fish stock assessment
Template Model Builder (TMB) is an R-package that is well suited for performing fast inference with latent Gaussian models when the likelihood can be written as a three times differentiable function. TMB automatically differentiates the likelihood and utilizes Markov structures to efficiently integrate over latent variables with the Laplace approximation. In this seminar I will first introduce TMB and then elaborate a fish stock assessment model implemented with use of TMB that provides quota advice for approximately 25 different fish species in Europe.
Published Sep. 2, 2021 2:10 PM
- Last modified Sep. 6, 2021 11:21 AM