Gradient-based Single objective optimizers

Gradient-Based optimizers use the gradient of an objective function to find a local minimum in the search space. If a function is not convex this local optimum might not be the global one.

Collaboration diagram for Gradient-based Single objective optimizers:

Classes
class	shark::AbstractLineSearchOptimizer< SearchPointType >
	Basis class for line search methods. More...
class	shark::Adam< SearchPointType >
	Adaptive Moment Estimation Algorithm (ADAM) More...
class	shark::BFGS< SearchPointType >
	Broyden, Fletcher, Goldfarb, Shannon algorithm for unconstraint optimization. More...
class	shark::CG< SearchPointType >
	Conjugate-gradient method for unconstrained optimization. More...
class	shark::LBFGS< SearchPointType >
	Limited-Memory Broyden, Fletcher, Goldfarb, Shannon algorithm. More...
class	shark::LineSearch< SearchPointType >
	Wrapper for the linesearch class of functions in the linear algebra library. More...
class	shark::Rprop< SearchPointType >
	This class offers methods for the usage of the Resilient-Backpropagation-algorithm with/out weight-backtracking. More...
class	shark::SteepestDescent< SearchPointType >
	Standard steepest descent. More...
class	shark::TrustRegionNewton
	Simple Trust-Region method based on the full Hessian matrix. More...