AD Model Builder

Summary:
AD Model Builder (ADMB) was specifically designed for complex highly-parameterized nonlinear models. ADMB uses automatic differentiation to provide the function optimizer with exact derivatives.

URL: http://admb-project.org/

Developers:

• Dave Fournier

Mode:	Reverse
Method:	Operator overloading
Supported Language:	C/C++

Reference:


Features:
ADMB plays a major role as a software environment for fisheries stock assessment. For example, the ADMB based general model Stock Synthesis (Methot 2009) is used to conduct the assessments of most major fisheries assessments in the United States and is also used in many other countries. ADMB also includes a Bayesian MCMC algorithm, likelihood profile calculations, simulation capabilities, and extensive matrix algebra functionality. A recent extension of ADMB adds the capability of modeling random efects, using the Laplace approximation or importance sampling (Skaug and Fournier 2006). ADMB is based on C++ coding (using the AUTODIF library) with additional structure such as keywords, functions, and operators. Knowledge of C++ is not a prerequisite for a successful ADMB user, because a simplified coding environment shields a user from advanced C++ concepts.

ADMB includes the AUTODIF automatic differentiation C++ library as open source.

Awards

Dave Fournier received the American Fisheries Society’s Ricker award. This is a prestigious award and provides Dave recognition not only for the great contribution AD Model Builder has made to fisheries stock assessment, but also Dave’s contribution to fisheries stock assessment methodology, such as being the pioneer of integrated analysis.

Supported Platforms:

• Windows
• Unix/Linux
• Mac

Licensing: open source

Entries in our publication database that actually use AD Model Builder in the numerical experiments:  7

The following diagram shows these entries versus the year of the publication.

10+ #Entries 0	1	1	1	1	2	1
	'03	'04	'06	'08	'09	'12
	Year