Introduction to maximum likelihood using ADMB

The course reviews probability theory and likelihood concepts, and covers maximum likelihood and other likelihood based applications using AD Model Builder (ADMB).  ADMB has been widely used for fitting nonlinear and non-normal statistical models, particularly fishery stock assessments, since the early 1990s.  Applications will focus on fixed effect models and penalized likelihood, although there will be brief coverage of fitting models by maximum likelihood in the presence of random effects, and limitations of the software in this context relative to Template Model Builder (TMB).  Developing models in TMB requires some knowledge of C++ and the needed basics of C++ coding will be covered.  This course is intended for those who know they need to use ADMB because of legacy code or current agency use. Others should consider the TMB course.  Students will need to install software used in the course prior to starting the course.  

Instructor: Dr. James Bence

Course Format and Sections

The class is asynchronous and online and can be started anytime.

Technology requirements

The course will be taught using R Studio, R, ADMB, and emacs (using admb modes).  This software is all freely available and instructions/assistance for installation will be provided prior to class to enrolled students.

Purchasing the class

You can purchase the class using a credit card or ACH at the QFC Storefront

MSU Guest Account (for non-MSU affiliated students)

Every student in the class needs an MSU account.  If you are not affiliated with MSU then you can get an MSU Guest Account here.

For questions, to pay by check, or to purchase classes in bulk contact Charlie Belinsky at 517-355-0126 or