Abstract
Developments in the theory and practice of unconstrained optimization are described. Both line search and trust region prototypes are explained and various algorithms based on the use of a quadratic model are outlined. The properties and implementation of the BFGS method are described in some detail and the current preference for this approach is discussed. Various conjugate gradient methods for large unstructured systems are given, but it is argued that limited memory methods are considerably more effective without a great increase in storage or time. Further developments in this area are described. For structured problems the possibilities for updates that retain sparsity are described, including a recent proposal which maintains positive definite matrices and reduces to the BFGS update in the dense case. The alternative use of structure in partially separable optimization is also discussed
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Al-Baali M. (1985) Descent property and global convergence of the Fletcher-Reeves method with inexact line search, J. Inst. Maths. Applns., 5, 121–124
Allwright J.C. (1988) Positive semi-definite matrices: Characterization via conical hulls and least-squares solution of a matrix equation, SIAM J. Control Optim., 26, 537–556 (see also errata in SIAM J. Control Optim., 28, (1990), 250–251)
Barzilai J. and Borwein J.M. (1988) Two-point step size gradient methods, IMA J. Numer. Anal., 8, 141–148
Beale E.M.L. (1972) A derivation of conjugate gradients, in Numerical Methods for Nonlinear Optimization, ed. F.A.Lootsma, Academic Press, London
Broyden C.G. (1967) Quasi-Newton methods and their application to function minimization, Maths, of Computation, 21, 368–381
Buckley A. and LeNir A. (1983) QN-like variable storage conjugate gradients, Math. Programming, 27, 155–175
Byrd R.H., Liu D.C. and Nocedal J. (1990) On the behavior of Broyden’s class of quasi-Newton methods, Report NAM-01, Northwestern Univ., Dept. EECS
Byrd R.H. and Nocedal J. (1989) A tool for the analysis of quasi-Newton methods with application to unconstrained minimization, SIAM J. Numer. Anal., 26, 727–739
Byrd R.H., Nocedal J. and Yuan Y. (1987) Global convergence of a class of quasi-Newton methods on convex problems, SIAM J. Numer. Anal., 24, 1171–1190
Conn A.R., Gould N.I.M. and Toint Ph.L. (1988) Testing a class of methods for solving minimization problems with simple bounds on the variables, Maths, of Computation, 50, 399–430
Conn A.R., Gould N.I.M. and Toint Ph.L. (1991) Convergence of quasi-Newton matrices generated by the symmetric rank one update, Math. Programming, 50, 177–195
Conn A.R., Gould N.I.M. and Toint Ph.L. (1992) LANCELOT: a Fortran package for large-scale nonlinear optimization (Release A), Springer Series in Computational Mathematics 17, Springer-Verlag, Heidelberg
Crowder H.P. and Wolfe P. (1972) Linear convergence of the conjugate gradient method, IBM J. Res. and Dev., 16, 431–433
Davidon W.C. (1959) Variable metric methods for minimization, Argonne Nat. Lab. Report ANL-5990 (rev.), and in SIAM J. Optimization, 1, (1991), 1–17
Davidon W.C. (1975) Optimally conditioned optimization algorithms without line searches, Math. Programming, 9, 1–30
Dembo R.S., Eisenstat S.C. and Steihaug T. (1982) Inexact Newton methods, SIAM J. Numer. Anal., 19, 400–408
Dennis J.E. and Moré J.J. (1974) A characterization of superlinear convergence and its application to quasi-Newton methods, Maths, of Computation, 28, 549–560
Dennis J.E. and Schnabel R.B. (1983) Numerical Methods for Unconstrained Optimization ana Nonlinear Equations, Prentice-Hall Inc., Englewood Cliffs, NJ
Dixon L.C.W. (1972) Quasi-Newton algorithms generate identical points, Math. Programming, 2, 383–387
Fletcher R. (1970) A new approach for variable metric algorithms, Computer J., 13, 317–322
Fletcher R. (1987) Practical Methods of Optimization, (2nd. Edition), John Wiley, Chichester
Fletcher R. (1990) Low storage methods for unconstrained optimization, in Computational Solution of Nonlinear Systems of Equations, eds. E.L.Allgower and K.Georg, Lectures in Applied Mathematics Vol. 26, AMS Publications, Providence, RI
Fletcher R. (1991) A new result for quasi-Newton formulae, SIAM J. Optimization, 1, 18–21
Fletcher R. (1992) An optimal positive definite update for sparse Hessian matrices, Report NA/145, Univ. of Dundee, to appear in SIAM J. Optimization
Fletcher R. and Powell M.J.D. (1963) A rapidly convergent descent method for minimization, Computer J., 3, 163–168
Fletcher R. and Reeves C.M. (1964) Function minimization by conjugate gradients, Computer J., 7, 149–154
Forsgren A., Gill P.E. and Murray W. (1993) Computing modified Newton directions using a partial Cholesky factorization, Royal Inst, of Tech., Sweden, Dept. of Math. Report TRITA/MAT-93–09
Gilbert J.C. and Nocedal J. (1990) Global convergence properties of conjugate gradient methods for optimization, Report 1268, INRIA
Gill P.E. and Murray W. (1972) Quasi-Newton methods for unconstrained optimization, J. Inst. Maths. Applns., 9, 91–108
Goldfarb D. (1970) A family of variable metric methods derived by variational means, Maths. of Computation, 24, 23–26
Goldfarb D. (1976) Factorized variable metric methods for unconstrained optimization, Maths. of Computation, 30, 796–811
Grandinetti (1979) Factorization versus nonfactorization in quasi-Newtonian algorithms for differentiable optimization, in Methods for Operations Research, Hain Verlay
Greenstadt J.L. (1970) Variations of variable metric methods, Maths. of Computation, 24, 1–22
Griewank A. and Toint Ph.L. (1982) Partitioned variable metric updates for large structured optimization problems, Numerische Math., 39, 429–448
Grippo L., Lampariello F. and Lucidi S. (1990) A quasi-discrete Newton algorithm with a nonmonotone stabilization technique, J. Optim. Theo. Applns., 64, 485–500
Hestenes M.R. and Stiefel E. (1952) Methods of conjugate gradients for solving linear systems, J. Res. NBS, 49, 409–436
Khalfan H., Byrd R.H. and Schnabel R.B. (1990) A theoretical and experimental study of the symmetric rank one update, Report CU-Cs-489–90, Univ. of Colorado at Boulder, Dept. CS
Lalee M. and Nocedal J. (1991) Automatic scaling strategies for quasi-Newton updates, Report NAM-04, Northwestern Univ., Dept. EECS
Lefebure B. (1993) Investigating new methods for large scale unconstrained optimization. M.Sc. Thesis, Univ. of Dundee, Dept. of Maths, and CS
Liu D.C. and Nocedal J. (1989) On the limited memory BFGS method for large scale optimization, Math. Programming, 45, 503–528
Luksan L. (1993) Computational experience with known variable metric updates, Working paper
Moré J.J. (1983) Recent developments in algorithms and software for trust region methods, in Mathematical Programming, The State of the Art, eds. A.Bachem, M Grotschel and G.Korte, Springer-Verlag, Berlin
Moré J.J., Garbow B.S. and Hillstrom K.E. (1981) Testing unconstrained optimization software, ACM Trans. Math. Softw., 7, 17–41
Moré J.J. and Sorensen D.C. (1979) On the use of directions of negative curvature in a modified Newton method, Math. Programming, 16, 1–20
Nash S.C. (1985) Preconditioning of truncated-Newton methods, SIAM J. S.S.C., 6, 599–618
Nash S.C. and Nocedal J. (1989) A numerical study of the limited memory BFGS method and the truncated-Newton method for large scale optimization, Report NAM-02, Northwestern Univ., Dept. EECS
Nocedal J. (1980) Updating quasi-Newton matrices with limited storage, Maths. of Computation, 35, 773–782
Nocedal J. and Yuan Y. (1991) Analysis of a self-scaling quasi-Newton method, Report NAM-02, Northwestern Univ., Dept. EECS
O’Leary D.P. (1982) A discrete Newton algorithm for minimizing a function of many variables, Math. Programming, 23, 20–33
Oren S.S. (1974) On the selection of parameters in self-scaling variable metric algorithms, Math. Programming, 7, 351–367
Oren S.S. and Luenberger D.G (1974) Self-scaling variable metric (SSVM) algorithms: I Criteria and sufficient conditions for scaling a class of algorithms, Management Sci., 20, 845–862
Oren S.S. and Spedicato E. (1976) Optimal conditioning of self-scaling variable metric algorithms, Math. Programming, 10, 70–90
Polak E. and Ribière G. (1969) Note sur la convergence de methodes de directions conjugées, Rev. Française Informat Recherche Opérationnelle, 3e Année, 16, 35–43
Powell M.J.D. (1970) A new algorithm for unconstrained optimization, in Nonlinear Programming, eds. J.B.Rosen, O.L.Mangasarian and K.Ritter, Academic Press, New York
Powell M.J.D. (1971) On the convergence of the variable metric algorithm, J. Inst. Maths. Applns., 7, 21–36
Powell M.J.D. (1976) Some global convergence properties of a variable metric algorithm for minimization without exact line searches, in Nonlinear Programming, SIAM-AMS Proceedings Vol. IX, eds. R.W.Cottle and C. E. Lemke, SIAM Publications, Philadelphia
Powell M.J.D. (1977) Restart procedures for the conjugate gradient method, Math. Programming, 12, 241–254
Powell M.J.D. (1984) Nonconvex minimization calculations and the conjugate gradient method, in Numerical Analysis, Dundee 1983, ed. D.F.Griffiths, Lecture Notes in Mathematics 1066, Springer Verlag, Berlin
Powell M.J.D. (1986) How bad are the BFGS and DFP methods when the objective function is quadratic?, Math. Programming, 34, 34–47
Powell M.J.D. (1987) Updating conjugate directions by the BFGS formula, Math. Programming, 38, 29–46
Powell M.J.D. and Toint Ph. L. (1979) On the estimation of sparse Hessian matrices, SIAM J. Numer. Anal., 16, 1060–1074
Schnabel R.B. (1983) Quasi-Newton methods using multiple secant equations, Report CU-CS-247–83, Univ. of Colorado at Boulder, Dept. CS
Schnabel R.B. and Eskow E. (1990) A new modified Cholesky factorization, SIAM J. Sci. Stat. Comp., 11, 1136–1158
Shanno D.F. (1970) Conditioning of quasi-Newton methods for function minimization, Maths. of Computation, 24, 647–656
Shanno D.F. and Phua K.H. (1980) Remark on algorithm 500, minimization of unconstrained multivariate functions, ACM Trans. Math. Softw., 6, 618–622
Siegel D. (1991) Updating of conjugate direction matrices using members of Broyden’s family, Rep. DAMTP 1991/NA4, Univ. of Cambridge, Dept. AMTP
Siegel D. (1991) Modifying the BFGS update by a new column scaling technique, Report DAMTP 1991/NA5, Univ. of Cambridge, Dept. AMTP
Siegel D. (1992) Implementing and modifying Broyden class updates for large scale optimization, Report DAMTP 1992/NA12, Univ. of Cambridge, Dept. AMTP
Sorensen D.C. (1982) Collinear scaling and sequential estimation in sparse optimization algorithms, Math. Programming Study, 18, 135–159
Steihaug T. (1983) The conjugate gradient method and trust regions in large scale optimization, SIAM J. Numer. Anal., 20, 626–637
Toint Ph. L. (1977) On sparse and symmetric updating subject to a linear equation. Maths, of Computation, 31, 954–961
Toint Ph. L. (1981) A note on sparsity exploiting quasi-Newton methods. Math. Programming, 21, 172–181
Toint Ph. L. (1981) Towards an efficient sparsity exploiting Newton method for minimization, in Sparse Matrices and Their Uses, ed. I.S.Duff, Academic Press, London
Torczon V. (1991) On the convergence of the multidimensional search algorithm, SIAM J. Optimization, 1, 123–145
Wood G.R. (1992) The bisection method in higher dimensions, Math. Programming, 55, 319–338
Zhang Y. and Tewarson R.P. (1988) Quasi-Newton algorithms with updates from the pre-convex part of Broyden’s family, IMA J. Numer. Anal., 8, 487–509
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1994 Kluwer Academic Publishers
About this chapter
Cite this chapter
Fletcher, R. (1994). An Overview of Unconstrained Optimization. In: Spedicato, E. (eds) Algorithms for Continuous Optimization. NATO ASI Series, vol 434. Springer, Dordrecht. https://doi.org/10.1007/978-94-009-0369-2_5
Download citation
DOI: https://doi.org/10.1007/978-94-009-0369-2_5
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-6652-5
Online ISBN: 978-94-009-0369-2
eBook Packages: Springer Book Archive