Norm descent conjugate gradient methods for solving symmetric nonlinear equations

摘要: Nonlinear conjugate gradient method is very popular in solving large-scale unconstrained minimization problems due to its simple iterative form and lower storage requirement. In the recent years, it was successfully extended solve higher-dimension monotone nonlinear equations. Nevertheless, research activities on symmetric equations are just beginning. This study aims developing, analyzing, validating a family of methods for The proposed algorithms based latest, state-of-the-art descent minimization. series derivative-free, where Jacobian information needless at full iteration process. We prove that converge globally under some appropriate conditions. Numerical results with differentiable parameter's values performance comparisons another solver CGD demonstrate superiority effectiveness reported.