L-shaped decomposition of two-stage stochastic programs with integer recourse

We consider two-stage stochastic programming problems with integer recourse. The L-shaped method of stochastic linear programming is generalized to these problems by using generalized Benders decomposition. Nonlinear feasibility and optimality cuts are determined via general duality theory and can be generated when the second stage problem is solved by standard techniques. Finite convergence of the method is established when Gomory’s fractional cutting plane algorithm or a branch-and-bound algorithm is applied.     

Parallel decomposition of multistage stochastic programming problems

A new decomposition method for multistage stochastic linear programming problems is proposed. A multistage stochastic problem is represented in a tree-like form and with each node of the decision tree a certain linear or quadratic subproblem is associated. The subproblems generate proposals for their successors and some backward information for their predecessors. The subproblems can be solved in parallel and exchange information in an asynchronous way through special buffers. After a finite time the method either finds an optimal solution to the problem or discovers its inconsistency. An analytical illustrative example shows that parallelization can speed up computation over every sequential method. Computational experiments indicate that for large problems we can obtain substantial gains in efficiency with moderate numbers of processors.This work was partly supported by the International Institute for Applied Systems Analysis, Laxenburg, Austria.     

A comparative study of decomposition algorithms for stochastic combinatorial optimization

This paper presents comparative computational results using three decomposition algorithms on a battery of instances drawn from two different applications. In order to preserve the commonalities among the algorithms in our experiments, we have designed a testbed which is used to study instances arising in server location under uncertainty and strategic supply chain planning under uncertainty. Insights related to alternative implementation issues leading to more efficient implementations, benchmarks for serial processing, and scalability of the methods are also presented. The computational experience demonstrates the promising potential of the disjunctive decomposition (D ²) approach towards solving several large-scale problem instances from the two application areas. Furthermore, the study shows that convergence of the D ² methods for stochastic combinatorial optimization (SCO) is in fact attainable since the methods scale well with the number of scenarios.     

Cutting plane algorithms for solving a stochastic edge-partition problem

We consider the edge-partition problem, which is a graph theoretic problem arising in the design of Synchronous Optical Networks. The deterministic edge-partition problem considers an undirected graph with weighted edges, and simultaneously assigns nodes and edges to subgraphs such that each edge appears in exactly one subgraph, and such that no edge is assigned to a subgraph unless both of its incident nodes are also assigned to that subgraph. Additionally, there are limitations on the number of nodes and on the sum of edge weights that can be assigned to each subgraph. In this paper, we consider a stochastic version of the edge-partition problem in which we assign nodes to subgraphs in a first stage, realize a set of edge weights from a finite set of alternatives, and then assign edges to subgraphs. We first prescribe a two-stage cutting plane approach with integer variables in both stages, and examine computational difficulties associated with the proposed cutting planes. As an alternative, we prescribe a hybrid integer programming/constraint programming algorithm capable of solving a suite of test instances within practical computational limits.     

Cross decomposition applied to the stochastic transportation problem

In this paper we give a solution method for the stochastic transportation problem based on Cross Decomposition developed by Van Roy (1980). Solution methods to the derived sub and master problems are discussed and computational results are given for a number of large scale test problems. We also compare the efficiency of the method with other methods suggested for the stochastic transportation problem: The Frank-Wolfe algorithm and separable programming.     

Multi-period stochastic portfolio optimization: Block-separable decomposition

We consider a multiperiod stochastic programming recourse model for stock portfolio optimization. The presence of various risk and policy constraints leads to significant period-by-period linkage in the model. Furthermore, the dimensionality of the model is large due to many securities under consideration. We propose exploiting block separable recourse structure as well as methods of inducing such structure within nested L-shaped decomposition. We test the model and solution methodology with a base consisting of the Standard & Poor 100 stocks and experiment with several variants of the block separable technique. These are then compared to the standard nested period-by-period decomposition algorithm. It turns out that for financial optimization models of the kind that are discussed in this paper, significant computational efficiencies can be gained with the proposed methodology.     

The complexity of branch-and-price algorithms for the capacitated vehicle routing problem with stochastic demands

The capacitated vehicle routing problem with stochastic demands (CVRPSD) is a variant of the deterministic capacitated vehicle routing problem where customer demands are random variables. While the most successful formulations for several deterministic vehicle-routing problem variants are based on a set-partitioning formulation, adapting such formulations for the CVRPSD under mild assumptions on the demands remains challenging. In this work we provide an explanation to such challenge, by proving that when demands are given as a finite set of scenarios, solving the LP relaxation of such formulation is strongly NP-Hard. We also prove a hardness result for the case of independent normal demands.     

On augmented Lagrangian decomposition methods for multistage stochastic programs

A general decomposition framework for large convex optimization problems based on augmented Lagrangians is described. The approach is then applied to multistage stochastic programming problems in two different ways: by decomposing the problem into scenarios and by decomposing it into nodes corresponding to stages. Theoretical convergence properties of the two approaches are derived and a computational illustration is presented.     

Mslip: A computer code for the multistage stochastic linear programming problem

This paper describes an efficient implementation of a nested decomposition algorithm for the multistage stochastic linear programming problem. Many of the computational tricks developed for deterministic staircase problems are adapted to the stochastic setting and their effect on computation times is investigated. The computer code supports an arbitrary number of time periods and various types of random structures for the input data. Numerical results compare the performance of the algorithm to MINOS 5.0.     

Iterative algorithms of domain decomposition for the solution of a quasilinear elliptic problem

This paper deals with iterative algorithms for domain decomposition applied to the solution of a quasilinear elliptic problem. Two iterative algorithms are examined: the first one is the Schwarz alternating procedure and the second algorithm is suitable for parallel computing. Convergence results are established in the two-domain and multidomain decomposition cases. Some issues of parallel implementation of these algorithms are discussed.     

Efficient decomposition and linearization methods for the stochastic transportation problem

The stochastic transportation problem can be formulated as a convex transportation problem with nonlinear objective function and linear constraints. We compare several different methods based on decomposition techniques and linearization techniques for this problem, trying to find the most efficient method or combination of methods. We discuss and test a separable programming approach, the Frank-Wolfe method with and without modifications, the new technique of mean value cross decomposition and the more well known Lagrangean relaxation with subgradient optimization, as well as combinations of these approaches. Computational tests are presented, indicating that some new combination methods are quite efficient for large scale problems.     

On the convergence of stochastic dual dynamic programming and related methods

We discuss the almost-sure convergence of a broad class of sampling algorithms for multistage stochastic linear programs. We provide a convergence proof based on the finiteness of the set of distinct cut coefficients. This differs from existing published proofs in that it does not require a restrictive assumption.     

Extremization of multi-objective stochastic fractional programming problem

This paper addresses classes of assembled printed circuit boards, which faces certain kinds of errors during its process of manufacturing. Occurrence of errors may lead the manufacturer to be in loss. The encountered problem has two objective functions, one is fractional and the other is a non-linear objective. The manufacturers are confined to maximize the fractional objective and to minimize the non-linear objective subject to stochastic and non-stochastic environment. This problem is decomposed into two problems. A solution approach to this model has been developed in this paper. Results of some test problems are provided.     

Finite master programs in regularized stochastic decomposition

Stochastic decomposition is a stochastic analog of Benders' decomposition in which randomly generated observations of random variables are used to construct statistical estimates of supports of the objective function. In contrast to deterministic Benders' decomposition for two stage stochastic programs, the stochastic version requires infinitely many inequalities to ensure convergence. We show that asymptotic optimality can be achieved with a finite master program provided that a quadratic regularizing term is included. Our computational results suggest that the elimination of the cutting planes impacts neither the number of iterations required nor the statistical properties of the terminal solution.This work was supported in part by Grant No. AFOSR-88-0076 from the Air Force Office of Scientific Research and Grant Nos. DDM-89-10046, DDM-9114352 from the National Science Foundation.Corresponding author.     

The traveling purchaser problem with stochastic prices: Exact and approximate algorithms

The paper formulates an extension of the traveling purchaser problem where multiple types of commodities are sold at spatially distributed locations with stochastic prices (each following a known probability distribution). A purchaser’s goal is to find the optimal routing and purchasing strategies that minimize the expected total travel and purchasing costs needed to purchase one unit of each commodity. The purchaser reveals the actual commodity price at a seller upon arrival, and then either purchases the commodity at the offered price, or rejects the price and visits a next seller. In this paper, we propose an exact solution algorithm based on dynamic programming, an iterative approximate algorithm that yields bounds for the minimum total expected cost, and a greedy heuristic for fast solutions to large-scale applications. We analyze the characteristics of the problem and test the computational performance of the proposed algorithms. The numerical results show that the approximate and heuristic algorithms yield near-optimum strategies and very good estimates of the minimum total cost.     

Improving constants of strong convexity in linear stochastic programming

We derive formulas for constants of strong convexity (CSCs) of expectation functions encountered in two-stage stochastic programs with linear recourse. One of them yields a CSC as the optimal value of a certain quadratically constrained quadratic program, another one in terms of the thickness of the feasibility polytope of the dual problem associated to the recourse problem. CSCs appear in Hoelder-type estimates relating the distance of optimal solution sets of stochastic programs to a suitable distance of underlying probability distributions.     

The stochastic trim-loss problem

The cutting stock problem (CSP) is one of the most fascinating problems in operations research. The problem aims at determining the optimal plan to cut a number of parts of various length from an inventory of standard-size material so to satisfy the customers demands. The deterministic CSP ignores the uncertain nature of the demands thus typically providing recommendations that may result in overproduction or in profit loss. This paper proposes a stochastic version of the CSP which explicitly takes into account uncertainty. Using a scenario-based approach, we develop a two-stage stochastic programming formulation. The highly non-convex nature of the model together with its huge size prevent the application of standard software. We use a solution approach designed to exploit the specific problem structure. Encouraging preliminary computational results are provided.     

Heuristic and lower bound for a stochastic location-routing problem

In this article a stochastic location-routing problem is defined and cast as a two-stage model. In a first stage the set of plants and a family of routes are determined; in a second stage a recourse action is applied to adapt these routes to the actual set of customers to visit, once they are known. A two-phase heuristic is developed. An initial feasible solution is built by solving a sequence of subproblems, and an improvement phase is then applied. A lower bound based on bounding separately different parts of the cost of any feasible solution is also developed. Computational results are reported.     

Stochastic semidefinite programming: a new paradigm for stochastic optimization

Semidefinite programs are a class of optimization problems that have been studied extensively during the past 15 years. Semidefinite programs are naturally related to linear programs, and both are defined using deterministic data. Stochastic programs were introduced in the 1950s as a paradigm for dealing with uncertainty in data defining linear programs. In this paper, we introduce stochastic semidefinite programs as a paradigm for dealing with uncertainty in data defining semidefinite programs.The work of this author was supported in part by the U.S. Army Research Office under Grant DAAD 19-00-1-0465. The material in this paper is part of the doctoral dissertation of this author in preparation at Washington State University.