Table of Contents
AdvancedMiner offers an interface for using the Opt++ library of nonlinear optimization algorithms developed in Sandia National Laboratories, USA (web page: http://csmr.ca.sandia.gov/). The library includes classic Newton methods as well as conjugate gradient and bound constrained ellipsoid methods.
The general form of the optimization problem may be stated as follows
In this problem, the objective function
is assumed to be a general nonlinear function (NLF). The constraint functions
and 
may also be
general nonlinear functions, however, algorithms which are part of
AdvancedMiner do not work in such a general setting.
Currently only algorithms dealing with linear constraint functions are
available.
In order to solve an optimization problem one should choose an
adequate algorithm depending on the objective function
and the existence
of constraints.
If the adequate optimization algorithm uses for example the gradient
of the objective function 
, then the function 
should be an object with
an evaluate method, which allows to evaluate the gradient.