In this paper, we propose a modified scaling BFGS method for unconstrained minimization.A remarkable feature of the proposed method is that it can improve the performance of the BFGS method and possesses a global convergence property without convexity assumption on the objective function.Under certain assumptions, we also establish superlinear convergence of the method.Finally we show numerical results.
Sugasawa et al. (Tue,) studied this question.