Algorithms for Distributionally Risk-Receptive and Robust Stochastic Integer Programs and Interdiction Problems
| dc.contributor.author | Kang, Sumin | en |
| dc.contributor.committeechair | Bansal, Manish | en |
| dc.contributor.committeemember | Sarin, Subhash C. | en |
| dc.contributor.committeemember | Chen, Xi | en |
| dc.contributor.committeemember | Tunc, Sait | en |
| dc.contributor.department | Industrial and Systems Engineering | en |
| dc.date.accessioned | 2025-08-15T08:00:18Z | en |
| dc.date.available | 2025-08-15T08:00:18Z | en |
| dc.date.issued | 2025-08-14 | en |
| dc.description.abstract | This dissertation advances the theory and algorithms of stochastic mixed-integer programs under distributional ambiguity, with a particular focus on adversarial games and interdiction problems. We introduce models within both distributionally risk-receptive (DRR) and distributionally robust optimization (DRO) frameworks, allowing for an adjustment of decision-maker's risk attitudes in adversarial settings. A key contribution is the study of an optimistic perspective, modeled by the DRR framework, which complements a pessimistic perspective of the conventional DRO paradigm and offers new insights for analyzing adversarial models. This research considers many algebraic optimization modeling tools such as two-stage and multi-stage stochastic programs, integer and disjunctive programming, decision-independent and decision-dependent uncertainty, and both single- and bi-parameterized recourse problems, in the DRR and DRO frameworks. In particular, the dissertation explores three main directions. In the first direction, we study DRR and DRO two-stage network interdiction problems. For solving them, we propose a new class of cutting planes and develop exact and approximation algorithms with finite convergence results. In the second direction, we extend these frameworks to multi-stage stochastic integer programs with both decision-independent and decision-dependent uncertainty, and introduce solution methods based on cutting plane and reformulation techniques. We further extend these results to multi-stage stochastic disjunctive programs by introducing an extended formulation for their feasible sets. These models are demonstrated on multi-period interdiction problems for network vulnerability assessment. In the third direction, we study two-stage stochastic integer programs with bi-parameterized recourse, where both the recourse objective and constraints are parameterized by first-stage decisions, under both min-min and min-max structures. These models arise in general interdiction problems, decision-dependent stochastic programs, and DRO problems with decision-dependent ambiguity sets that may be empty for some first-stage decisions. To solve these problems, we propose Lagrangian-integrated L-shaped (L^2) methods, including both exact and approximation variants. Our numerical experiments demonstrate that the proposed methods outperform existing approaches, achieving significant speedups in solving interdiction problem instances and general stochastic programming problem instances. Furthermore, through out-of-sample tests, we show that the DRR framework can effectively identify network vulnerabilities and provide robust solutions under data contamination. | en |
| dc.description.abstractgeneral | This dissertation advances the study of mathematical optimization tools for decision-making under uncertainty, with a particular focus on problems in adversarial settings, such as assessing infrastructure vulnerabilities under disruption. We model optimization problems with both risk-receptive and risk-averse decision-makers when the complete information of distributions are unavailable. These modeling frameworks are named distributionally risk-receptive (DRR) and distributionally robust optimization (DRO) frameworks, respectively. A key contribution is the study of an optimistic perspective, modeled by the DRR framework, which complements a pessimistic perspective of the conventional DRO paradigm and offers new insights for analyzing problems within adversarial environments. This research addresses a broad range of problem classes. Our settings allow decisions over multiple stages and include discrete or logically disjunctive variables (integer/disjunctive programming). We also consider cases where the realizations of uncertainty are either decision-dependent or decision-independent. For these problem classes, we develop exact and approximate solution methods. Our numerical experiments demonstrate that the proposed methods outperform existing approaches, achieving significant speedups. In terms of practical implications, each chapter of the dissertation explores a distinct direction. In the first direction, we study adversarial problems, namely network interdiction problems, where attacker and defender make decisions sequentially, modeled within the DRR and DRO frameworks. Solving the DRR model, especially, provides a pessimistic viewpoint of the defender, which can be applied to vulnerability analysis in real-world applications, such as critical infrastructure contingency planning. We show that such insights may be not attainable with conventional stochastic programming models, including DRO models. In the second direction, we extend the DRR and DRO frameworks to multistage decision making, where players act sequentially across multiple periods. These models are applied to multi-period interdiction problems. We demonstrate these models can identify vulnerable network components and determine robust allocation of security resources. We further show that the DRR models are effective in data corruption settings--where input data are adversarially corrupted--outperforming both risk-neutral and DRO models. In the third direction, we study two-stage optimization problems where first-stage decisions affect both the second-stage objective and constraints (i.e., bi-parameterized recourse). These models have broad applicability, including general interdiction problems, decision-dependent stochastic programs, and DRO problems with decision-dependent ambiguity sets. Notably, this framework extends conventional stochastic programming by capturing complex settings--common in real-world applications like supply chain management and network security--where early decisions affect not only future costs but also the constraints governing subsequent actions. The bi-parameterized recourse approach thus enhances modeling fidelity for such complex decision-making environments. | en |
| dc.description.degree | Doctor of Philosophy | en |
| dc.format.medium | ETD | en |
| dc.identifier.other | vt_gsexam:44460 | en |
| dc.identifier.uri | https://hdl.handle.net/10919/137507 | en |
| dc.language.iso | en | en |
| dc.publisher | Virginia Tech | en |
| dc.rights | In Copyright | en |
| dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | en |
| dc.subject | Distributionally Risk-Receptive Optimization | en |
| dc.subject | Distributionally Robust Optimization | en |
| dc.subject | Interdiction Problem | en |
| dc.subject | Stochastic Integer Programming | en |
| dc.subject | Decomposition Method | en |
| dc.title | Algorithms for Distributionally Risk-Receptive and Robust Stochastic Integer Programs and Interdiction Problems | en |
| dc.type | Dissertation | en |
| thesis.degree.discipline | Industrial and Systems Engineering | en |
| thesis.degree.grantor | Virginia Polytechnic Institute and State University | en |
| thesis.degree.level | doctoral | en |
| thesis.degree.name | Doctor of Philosophy | en |
Files
Original bundle
1 - 1 of 1