J. Linderoth and S.J. Wright. Decomposition algorithms for stochastic programming on a computational grid. Comput. Optim. Appl., 24(23) :207--250, 2003. Stochastic programming. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Kestrel: An Interface from Modeling Systems to the NEOS.. - Dolan, Fourer, Goux, Munson (Correct)
....for all of the subproblems to be solved before going on to the next master problem anyway. Other iterative schemes that generate numerous subproblems have a similar property. A more sophisticated Kestrel interface would be necessary, however, to deal with asynchronous decomposition schemes [22] that may solve a new master problem before all of the relevant subproblems have been received. When our script is run with the very small sample data file of Appendix B, the overhead of submitting and retrieving Kestrel jobs dominates the total elapsed time. For harder subproblems, however, we ....
J. Linderoth and S. Wright, Decomposition Algorithms for Stochastic Programming on a Computational Grid. Preprint ANL/MCS-P875-0401, Mathematics and Computer Science Division, Argonne National Laboratory (2001).
The Empirical Behavior of Sampling Methods for.. - Linderoth, Shapiro.. (2002) (2 citations) Self-citation (Linderoth Wright) (Correct)
No context found.
J. T. Linderoth and S. J. Wright. Decomposition algorithms for stochastic programming on a computational grid. Preprint ANL/MCS-P875-0401, Mathematics and Computer Science Division, Argonne National Laboratory, Argonne, Ill., April 2001. Available at http://www.optimization-online.org/DB_HTML/2001/04/315.html.
The Empirical Behavior of Sampling Methods for.. - Linderoth, Shapiro.. (2002) (2 citations) Self-citation (Linderoth Wright) (Correct)
....tests for optimality or near optimality of a candi 4 Je# Linderoth et al. date solution in Section 5. In Section 6 we focus on the two stage stochastic linear programming problem with recourse (2) outlining the software that we used to solve this problem. A fuller description can be found in [18]. Section 7 briefly highlights the characteristics of the computational grid we used in this work. Section 8 describes the benchmark problems used in our computational tests, while Section 9 describes and analyzes the results obtained from our tests. 2. Sample Average Approximations Suppose that ....
....Recourse In this section, we focus our attention on two stage stochastic linear programs with recourse over a discrete scenario space, which are the subjects of our computational experiments. We discuss briefly the algorithm used in the experiments, referring the reader to Linderoth and Wright [18] for further details. The problem we consider is (2) with second stage problems defined by (3) with fixed recourse, and with a finite number K of scenarios. We thus have defined by (4) where # k defines the data vector (q k , h k , T k , W ) k = 1, 2, K. The optimality conditions ....
[Article contains additional citation context not shown here]
J. T. Linderoth and S. J. Wright. Decomposition algorithms for stochastic programming on a computational grid. Preprint ANL/MCS-P875-0401, Mathematics and Computer Science Division, Argonne National Laboratory, Argonne, Ill., April 2001. Available at http://www.optimization-online.org/DB_HTML/2001/04/315.html.
Near-Optimal Adaptive Control of a Large Grid Application - Buaklee, Tracy, Vernon.. Self-citation (Wright) (Correct)
....Performance, Experimentation. Keywords Parallel algorithms, parallel application performance, stochastic optimization, adaptive computations, grid computing. 1. INTRODUCTION This paper develops a model for near optimal adaptive control of the state of the art stochastic optimization code ATR [18] on Grid platforms such as Condor [19] Globus [10] or Legion [13] in which the number and capabilities of the distributed hosts that execute ATR varies during the course of the computation. Stochastic optimization uses large amounts of computational resources to solve key organizational, ....
....of the algorithm (such as the time to initialize each iteration and the number of iterations to reach convergence) in ways that are not easily quantified. Furthermore, the runtime environment typically includes support functions that add unpredictable delays to task execution times. Previous work [18] in developing the ATR algorithm for Condor platforms has relied on simple task scheduling and extensive experimental measurements of total application running time as a function of (1) the average number of allocated compute nodes, and (2) the fixed (i.e. non adaptive) values of the ATR ....
[Article contains additional citation context not shown here]
Linderoth, J. and Wright, S. J., "Decomposition Algorithms for Stochastic Programming on a Computational Grid," Preprint ANIfMCS-P875-0401, MCS Division, Argonne National Laboratory, April, 2001.
Computational Grids for Stochastic Programming - Linderoth, Wright Self-citation (Linderoth Wright) (Correct)
....We now discuss algorithms for two stage stochastic linear programming with recourse and their implementation on the grid platform described in the previous section. A more complete discussion of these algorithms, their implementations, and computational results obtained with them can be found in [20]. Our target problem has the following form: min x Q(x) def = c T x N # i=1 p i Q i (x) subject to Ax = b, x # 0, 1) where each Q i (x) is the value function for a second stage linear program: Q i (x) def = min y(# i ) q(# i ) T y(# i ) subject to (2a) Wy(# i ) h(# i ....
....Algorithms ALS, TR, and ATR 3.4. Computational Results We present a selection of results obtained with the solvers discussed in this section. For a fuller picture of our computational experience, in particular the dependence of the results on various parameter choices, we refer the reader to [20]. Table 1 shows results obtained from a sampled instance of SSN, a problem in telecommunications design described in [29] There are 89 first stage variables and a single constraint (a budget constraint) while the second stage problems have 706 unknowns, 175 constraints, and a constraint matrix ....
J. T. Linderoth and S. J. Wright. Decomposition algorithms for stochastic programming on a computational grid. Preprint ANL/MCS-P875-0401, Mathematics and Computer Science Division, Argonne National Laboratory, Argonne, Ill., April
Multi-Stage Investment Decision under Contingent Demand for - Networking Planning Miguel (Correct)
No context found.
J. Linderoth and S.J. Wright. Decomposition algorithms for stochastic programming on a computational grid. Comput. Optim. Appl., 24(23) :207--250, 2003. Stochastic programming.
Mean-Risk Objectives in Stochastic Programming - Ahmed (2004) (Correct)
No context found.
J. Linderoth and S. Wright. Decomposition algorithms for stochastic programming on a computational grid. Computational Optimization and Applications, 24:207--250, 2003.
Kestrel: An Interface from Modeling Systems to the NEOS.. - Dolan, Fourer, Goux, Munson (Correct)
No context found.
J. Linderoth and S. Wright, Decomposition Algorithms for Stochastic Programming on a Computational Grid. Preprint ANL/MCS-P875-0401, Mathematics and Computer Science Division, Argonne National Laboratory (2001).
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC