Subject Area
Industrial/Manufacturing Engineering
Abstract
In this dissertation, we present novel sampling-based algorithms for solving two-stage stochastic programming problems. Sampling-based methods provide an efficient approach to solving large-scale stochastic programs where uncertainty is possibly defined on continuous support. When sampling-based methods are employed, the process is usually viewed in two steps - sampling and optimization. When these two steps are performed in sequence, the overall process can be computationally very expensive. In this dissertation, we utilize the framework of internal-sampling where sampling and optimization steps are performed concurrently. The dissertation comprises of two parts. In the first part, we design a new sampling technique for solving two-stage stochastic linear programs with continuous recourse. We incorporate this technique within an internal-sampling framework of stochastic decomposition. In the second part of the dissertation, we design an internal-sampling-based algorithm for solving two-stage stochastic mixed-integer programs with continuous recourse. We design a new stochastic branch-and-cut procedure for solving this class of optimization problems. Finally, we show the efficiency of this method for solving large-scale practical problems arising in logistics and finance.
Degree Date
Fall 12-18-2021
Document Type
Thesis
Degree Name
Ph.D.
Department
Operations Research and Engineering Management
Advisor
Harsha Gangammanavar
Second Advisor
Halit Üster
Number of Pages
149
Format
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
Recommended Citation
Tabrizian, Siavash, "Sampling-Based Algorithms for Two-Stage Stochastic Programs and Applications" (2021). Operations Research and Engineering Management Theses and Dissertations. 18.
https://scholar.smu.edu/engineering_managment_etds/18
