A new three-term conjugate gradient algorithm for unconstrained optimization (Q120735)

The authors present a new three-term conjugate gradient algorithm for unconstrained optimization by minimization of the one-parametric quadratic model of the objective function. Under certain assumptions, the global convergence analysis of the proposed algorithm for uniformly convex functions is established. Some preliminary numerical results on a collection of 800 large-scale unconstrained optimization test problems are provided to compare the proposed algorithm with other existing ones and show that the considered algorithm is indeed more efficient and more robust.