Global Convergence of a Nonmonotone Trust Region Algorithm with Memory for Unconstrained Optimization
- 80 Downloads
In this paper, we consider a trust region algorithm for unconstrained optimization problems. Unlike the traditional memoryless trust region methods, our trust region model includes memory of the past iteration, which makes the algorithm less myopic in the sense that its behavior is not completely dominated by the local nature of the objective function, but rather by a more global view. The global convergence is established by using a nonmonotone technique. The numerical tests are also given to show the efficiency of our proposed method.
KeywordsTrust region Memory model Nonmonotone technique Global convergence
Mathematics Subject Classifications (2010)65K05 90C30
Unable to display preview. Download preview PDF.
- 1.Powell, M.J.D.: In: Mangasarian, O.L.M., Meyer, R.R., Robinson, S.M. (eds.) Convergence properties of a class minimization algorithms in nonlinear programming, vol. 2, pp. 47–67 (1975)Google Scholar
- 3.Yuan, Y.X.: Nonlinear Optimization: Trust Region Algorithms. Research Report, Institute of Computational Mathematics, Chinese Academy of Sciences (1999)Google Scholar
- 4.Sorensen, D.C.: Trust region methods for unconstrained optimization. In: Powell, M.J.D. (ed.) Nonlinear Optimization 1981, pp. 29–38. Academic, London (1982)Google Scholar
- 5.Conn, A.R., Gould, N.I.M., Toint, P.L.: Trust Region Methods. MSP-SIAM Series on Optimization. Philadelphia, PA (2000)Google Scholar
- 6.Gould, N.I.M., Lucidi, S., Roma, M., Toint, P.L.: A line search algorithm with memory for unconstrained optimization. In: De leone, R., Mauli, A., Pardalos, P., Toraldo, G. (eds.) High Performance Algorithms and Software in Nonlinear Optimization, pp. 207–223. Kluwer Academic Publishers (1998)Google Scholar
- 9.Ke, X.W., Han, J.Y.: A class of nonmonotone trust region algorithm for unconstrained optimization. Sci. China 28, 488–492 (1998)Google Scholar