A modified BFGS method and its superlinear convergence in nonconvex minimization with general line search rule (Q949372): Difference between revisions

The authors make efforts to establish a new algorithm for solving unconstrained optimization problems. It bases on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) method with a suitable line-search. It differs from the standard approach by making use of an auxiliary function which has the same global minimum as the objective function -- more precisely it depends on the minimizer. A global convergence analysis is carried out and shows a super-linear convergence. However, it remains obscure how an implementation could look like and the English is dissatisfying.