Ensemble Variance Reduction Methods for Stochastic Mixed-Integer Programming and their Application to the Stochastic Facility Location Problem

被引:6
|
作者
Xu, Jiajun [1 ]
Sen, Suvrajeet [2 ]
机构
[1] Univ Southern Calif, Ming Hsieh Dept Elect & Comp Engn, Los Angeles, CA 90089 USA
[2] Univ Southern Calif, Daniel J Epstein Dept Ind & Syst Engn, Los Angeles, CA 90089 USA
关键词
stochastic mixed-integer programming; sample average approximation; variance reduction; bagging; compromise decisions; facility location problem; SIMULATION BUDGET ALLOCATION;
D O I
10.1287/ijoc.2021.0324
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Sample average approximation (SAA), the standard approach to stochastic mixed-integer programming, does not provide guidance for cases with limited computational budgets. In such settings, variance reduction is critical in identifying good decisions. This paper explores two closely related ensemble methods to determine effective decisions with a probabilistic guarantee. (a) The first approach recommends a decision by coordinating aggregation in the space of decisions as well as aggregation of objective values. This combination of aggregation methods generalizes the bagging method and the "compromise decision" of stochastic linear programming. Combining these concepts, we propose a stopping rule that provides an upper bound on the probability of early termination. (b) The second approach applies efficient computational budget allocation for objective function evaluation and contributes to identifying the best solution with a predicted lower bound on the probability of correct selection. It also reduces the variance of the upper bound estimate at optimality. Furthermore, it adaptively selects the evaluation sample size. Both approaches provide approximately optimal solutions even in cases with a huge number of scenarios, especially when scenarios are generated by using oracles/simulators. Finally, we demonstrate the effectiveness of these methods via extensive computational results for "megascale" (extremely large scale) stochastic facility location problems.
引用
收藏
页码:587 / 599
页数:14
相关论文
共 50 条
  • [1] Integer set reduction for stochastic mixed-integer programming
    Venkatachalam, Saravanan
    Ntaimo, Lewis
    COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2023, 85 (01) : 181 - 211
  • [2] Integer set reduction for stochastic mixed-integer programming
    Saravanan Venkatachalam
    Lewis Ntaimo
    Computational Optimization and Applications, 2023, 85 : 181 - 211
  • [3] Fenchel decomposition for stochastic mixed-integer programming
    Lewis Ntaimo
    Journal of Global Optimization, 2013, 55 : 141 - 163
  • [4] Mixed-integer value functions in stochastic programming
    Schultz, R
    COMBINATORIAL OPTIMIZATION - EUREKA, YOU SHRINK: PAPERS DEDICATED TO JACK EDMONDS, 2003, 2570 : 171 - 184
  • [5] Fenchel decomposition for stochastic mixed-integer programming
    Ntaimo, Lewis
    JOURNAL OF GLOBAL OPTIMIZATION, 2013, 55 (01) : 141 - 163
  • [6] Stochastic mixed-integer programming for a spare parts inventory management problem
    Johannsmann, Leonie M.
    Craparo, Emily M.
    Dieken, Thor L.
    Fugenschuh, Armin R.
    Seitner, Bjoern O.
    COMPUTERS & OPERATIONS RESEARCH, 2022, 138
  • [7] Monotonic bounds in multistage mixed-integer stochastic programming
    Maggioni F.
    Allevi E.
    Bertocchi M.
    Computational Management Science, 2016, 13 (3) : 423 - 457
  • [8] On the Glivenko-Cantelli problem in stochastic programming: Mixed-integer linear recourse
    Georg Ch. Pflug
    Andrzej Ruszczyński
    Rüdiger Schultz
    Mathematical Methods of Operations Research, 1998, 47 : 39 - 49
  • [9] A Stochastic Mixed-Integer Programming approach to the energy-technology management problem
    Stoyan, Stephen J.
    Dessouky, Maged M.
    COMPUTERS & INDUSTRIAL ENGINEERING, 2012, 63 (03) : 594 - 606
  • [10] On the Glivenko-Cantelli problem in stochastic programming: Mixed-integer linear recourse
    Pflug, GC
    Ruszczynski, A
    Schultz, R
    MATHEMATICAL METHODS OF OPERATIONS RESEARCH, 1998, 47 (01) : 39 - 49