Abstract
The special constraint structure and large dimension are characteristic for multistage stochastic optimization. This results from modeling future uncertainty via branching process or scenario tree. Most efficient algorithms for solving this type of problems use certain decomposition schemes, and often only a part of the whole set of scenarios is taken into account in order to make the problem tractable.
We propose a primal–dual method based on constraint aggregation, which constructs a sequence of iterates converging to a solution of the initial problem. At each iteration, however, only a reduced sub-problem with smaller number of aggregate constraints has to be solved. Number of aggregates and their composition are determined by the user, and the procedure for calculating aggregates can be parallelized. The method provides a posteriori estimates of the quality of the current solution approximation in terms of the objective function value and the residual.
Results of numerical tests for a portfolio allocation problem with quadratic utility function are presented.
Similar content being viewed by others
References
A. Antipin, Gradient and proximal control processes, in: Problems of Cybernetics, Russian Academy of Sciences (1992) pp. 32–66 (in Russian).
J.R. Birge, Decomposition and partitioning methods for multistage stochastic linear programs, Oper. Res. 33 (1985) 989–1007.
J.R. Birge, The relationship between the L-shaped method and dual basis factorization for stochastic linear programming, in: Numerical Techniques for Stochastic Optimization, Springer Ser. Comput. Math., Vol. 10 (1988) pp. 267–272.
J.R. Birge, Current trends in stochastic programming computation and applications, Technical report, Dept. of Industrial and Operational Engineering, University of Michigan, Ann Arbor, MI (1995).
J.R. Birge, C.J. Donohue, D.F. Holmes and O.G. Svintsitski, A parallel implementation of the nested decomposition algorithm for multistage stochastic linear programs, Math. Program. 75 (1996) 327–352.
J.R. Birge and A.J. King, Unbiased methods for sampled cuts in stochastic programming, Technical report, Dept. of Industrial and Operational Engineering, University of Michigan, Ann Arbor, MI (1995).
J.R. Birge and F. Louveaux, Introduction to Stochastic Programming (Springer, New York, 1997).
J.R. Birge and J.-B.Wets, Designing approximation schemes for stochastic optimization problems, in particular for stochastic programs with recourse, Math. Program. Study 2 (1986) 754–102.
B.J. Chun and S.M. Robinson, Scenario analysis via bundle decomposition, Ann. Oper. Res. 56 (1995 39–63.
G.B. Dantzig and P.W. Glynn, Parallel processors for planning under uncertainty, Ann. Oper. Res. 22 (1990) 1–21.
G.B. Dantzig and G. Infanger, Large-scale stochastic linear programs-importance sampling and Benders decomposition, in: Computational and Applied Mathematics, I. Algorithms and Theory, Sel. Rev. Pap. IMACS 13th World Congr., Dublin/Ireland, 1991 (1992) pp. 111–120.
M. Davidson, Proximal point mappings and constraint aggregation principle, Technical report, WP–96–102, IIASA, Laxenburg, Austria (1996).
M. Davidson, Primal-dual constraint aggregation method in multistage stochastic programming, Technical report, WP 97–26, University of Trier, Germany (1997).
M.A.H. Dempster, Stochastic Programming, The Institute of Mathematics and its Applications Conference Series (Academic Press, London, 1980).
M.A.H. Dempster and R.T. Thompson, EVPI-based importance sampling solution procedures for multistage stochastic linear programmes on parallel mimd architectures, Technical report, Judge Institute of Management Studies, University of Cambridge, UK (1996).
N.C.P. Edirisinghe, Bound-based approximations in stochastic programming, Ann. Oper. Res. 85 (1999) 103–127.
Yu. Ermoliev, A. Kryazhimskii and A. Ruszczynski, Constraint aggregation principle in convex optimization, Math. Program. 76 (1997) 353–372.
Yu. Ermoliev and R. Wets, eds., Numerical Techniques for Stochastic Optimization (Springer, Berlin, 1988).
K. Frauendorfer, Barycentric scenario trees in convex multistage stochastic programming, Math. Program. 75 (1996) 277–294.
H.I. Gassmann, MSLiP: A computer code for the multistage stochastic linear programming problem, Math. Program., Ser. A 47 (1990) 407–423.
J.L. Higle and S. Sen, Stochastic decomposition: An algorithm for two-stage linear programs with recourse, Math. Oper. Res. 16 (1991) 650–669.
J.L. Higle and S. Sen, Duality and statistical tests of optimality for two stage stochastic programs, Math. Program. 75 (1996) 257–276.
J.L. Higle and S. Sen, Stochastic Decomposition (Kluwer Academic, 1996).
P. Kall, Solution methods in stochastic programming, in: System Modeling and Optimization, Proceedings of the 16th IFIP-TC7 Conference, Compiegne, France, July 5–9, 1993. Lect. Notes Control Inf. Sci., Vol. 197 (Springer, London, (994) pp. 3–22.
P. Kall, A. Ruszczynski and K. Frauendorfer, Approximation techniques in stochastic programming, in: Numerical Techniques for Stochastic Optimization, Springer Ser. Comput. Math., Vol. 10 (1988) pp. 33–64.
K. Kiwiel, Methods of Descent for Nondifferentiable Optimization (Springer, Berlin, 1985).
C. Lemaréchal, A.S. Nemirovskij and Y. Nesterov, New variants of bundle methods, Math. Prog. 69 (1995) 111–147.
C. Lemaréchal, J.J. Strodiot and A. Bihain, On a bundle algorithm for non-smooth optimization, in: Nonlinear Programming 4 (Academic Press, New York, 1981) pp. 245–282.
F.V. Louveaux, Multistage stochastic programs with block-separable recourse, Math. Program. Study 28 (1986) 48–62.
H.M. Markowitz, Portfolio Selection: Efficient Diversification of Investments (Yale University Press, 1959).
R. Mifflin, A modification and an extension of Lemaréchal's algorithm for nonsmooth minimization, Math. Prog. Study 17 (1982) 77–90.
J.M. Mulvey and H. Vladimirou, Applying the progressive hedging algorithm to stochastic generalized networks, Ann. Oper. Res. 31 (1991) 399–424.
NAG C Library Manual, Mark 4, Numerical Algorithms Group Ltd., UK (1996).
A.S. Nemirovskij and D.B.Yudin, Problem Complexity andMethod Efficiency in Optimization (Wiley, New York, 1983).
M.-C. Noel and Y. Smeers, Nested decomposition of multistage nonlinear programs with recourse, Math. Program. 37 (1987) 131–152.
S.M. Robinson, Extended scenario analysis, Ann. Oper. Res. 31 (1991) 385–397.
R.T. Rockafellar, Convex Analysis (Princeton University Press, Princeton, NJ, 1973).
R.T. Rockafellar and R.J.-B. Wets, Scenarios and policy aggregation in optimization under uncertainty, Math. Oper. Res. 16 (1991) 119–147.
A. Ruszczynski, Parallel decomposition of multistage stochastic programming problems, Math. Program. Ser. A 58 (1993) 201–228.
H. Schramm and J. Zowe, A version of the bundle idea for minimizing a non-smooth function: conceptual idea, convergence analysis, numerical results, SIAM J. Opt. 2 (1992) 121–152.
A. Shapiro, Asymptotic behavior of optimal solutions in stochastic programming, Math. Oper. Res. 18 (1993) 829–845.
R.J.-B. Wets, Large scale linear programming techniques in stochastic programming, in: Numerical Techniques for Stochastic Optimization, Springer Ser. Comput. Math., Vol. 10 (1988) pp. 65–93.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Davidson, M. Primal—Dual Constraint Aggregation with Application to Stochastic Programming. Annals of Operations Research 99, 41–58 (2000). https://doi.org/10.1023/A:1019232731587
Issue Date:
DOI: https://doi.org/10.1023/A:1019232731587