A Three-Dimensional Subspace Algorithm Based on the Symmetry of the Approximation Model and WYL Conjugate Gradient Method

In this paper, a three-dimensional subspace method is proposed, in which the search direction generated by minimizing approximation model of objective function subspace. The not unique, and alternatives can be chosen between symmetric quadratic conic specific criteria. Moreover, idea WLY conjugate gradient applied to characterize change adjacent iteration points. strategy initial stepsize nonmonotone line are adopted, global convergence presented algorithm established under mild assumptions. numerical experiments, we use collection 80 unconstrained optimization test problems show competitive performance method.