×

Limited memory BFGS method based on a high-order tensor model. (English) Zbl 1309.90050

Summary: This paper is aimed to employ a modified quasi-Newton equation in the framework of the limited memory BFGS method to solve large-scale unconstrained optimization problems. The modified secant equation is derived by means of a forth order tensor model to improve the curvature information of the objective function. The global and local convergence properties of the modified LBFGS method, on uniformly convex problems are also studied. The numerical results indicate that the proposed limited memory method is superior to the standard LBFGS method.

MSC:

90C06 Large-scale problems in mathematical programming
90C53 Methods of quasi-Newton type

Software:

L-BFGS
Full Text: DOI

References:

[1] Andrei, N.: An unconstrained optimization test functions collection. Adv. Model. Optim. 10, 147-161 (2008) · Zbl 1161.90486
[2] Dolan, E.D., Moré, J.J.: Benchmarking optimization software with performance profiles. Math. Program. 91(2), 201-203 (2002) · Zbl 1049.90004 · doi:10.1007/s101070100263
[3] Fletcher, R.: Practical Methods of Optimization, 2nd edn. Wiley, New York (1987) · Zbl 0905.65002
[4] Liu, D., Nocedal, J.: On the limited memory BFGS method for large-scale optimization. Math. Program. 45, 503-528 (1989) · Zbl 0696.90048 · doi:10.1007/BF01589116
[5] Nocedal, J., Wright, S.J.: Numerical Optimization. Springer Series in Operations Research, 2nd edn. Springer, New York (2006) · Zbl 1001.65065
[6] Sun, W., Yuan, Y.-X.: Optimization Theory and Methods Nonlinear Programming. Springer Optimization and Its Applications. Springer, New York (2006) · Zbl 1129.90002
[7] Zhang, J.Z., Xu, C.X.: Properties and numerical performance of quasi-Newton methods with modified quasi-Newton equations. J. Comput. Appl. Math. 137, 269-278 (2001) · Zbl 1001.65065 · doi:10.1016/S0377-0427(00)00713-5
[8] Zoutendijk, G.; Abadie, J. (ed.), Nonlinear programming, computational methods, 37-86 (1970), Amsterdam · Zbl 0336.90057
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.