Solves unconstrained nonlinear optimization problems with simple bounds
on the variables.
The solver includes several of the most popular
search step methods for unconstrained optimization.
Implements the Newton method, the quasi-Newton BFGS and inverse
BFGS method, the quasi-Newton DFP and inverse DFP method, the Fletcher-Reeves
and Polak-Ribiere conjugate-gradient method, and the Fletcher conjugate
The Newton and the quasi-Newton
methods is using a subspace minimization technique to handle rank problem.
The equations system is solved using singular value decomposition,
i.e. only dense equation systems are solved.
Large, sparse systems is better solved with other nonlinear Tomlab solvers,
or with the conjugate gradient methods.
The quasi-Newton methods use safe
guarding techniques to avoid rank problem in the updated matrix.
The line search used is a modified version of
an algorithm by Fletcher: Practical Methods of Optimization, 1987.
Bound constraints are treated
similar as described in
Gill, Murray, Wright: Practical Optimization, Academic Press, 1982.