Sanghani Center Student Spotlight: Sumin Kang

Graphic is the poster “Distributionally Ambiguous Multistage Stochastic Integer and Disjunctive Programs: Applications to Sequential Two-player Interdiction Games,” presented at the INFORMS Annual Meeting 2023.

Ph.D. student Sumin Kang was drawn to the Department of Industrial Systems Engineering by its prestigious faculty with particular expertise in optimization for logistics problems with uncertainty. 

“As a logistics enthusiast, I found their focus aligned very well with my own interests,” said Kang, who is advised by Manish Bansal, core faculty at the Sanghani Center.  “And the interdisciplinary environment fostering a multi-faceted approach in problem solving at the Sanghani Center added to Virginia Tech’s appeal.”

Kang’s research interests lie in optimization under uncertainty, with a focus on network optimization problems and vulnerability analysis.

“Specifically I am interested in optimization problems with incomplete information about distributions of uncertain parameters,” he said. “These optimization problems find an application in the network interdiction problem. The network interdiction problem involves a game between two players, the interdictor and the network user, where the network user aims to minimize the objective value like traveling cost and security threat level, while the interdictor aims to maximize disruption of network. 

“Solving this problem is valuable for identifying network vulnerabilities, particularly in cases of unexpected disruptions,” said Kang.  “My proposed solution approaches consider the interdictor’s varying risk appetite towards unknown distributions.”


Kang’s interest in this research began as a master’s degree student in logistics at Korea Aerospace University, when he started to struggle with logistic optimization problems, he said, because despite the prevalence of real-world uncertainty, the literature mainly focused on deterministic cases due to their high complexity. This motivated him to address the gap and contribute to the domain of optimization with uncertainty for large-scale problems.

One of Kang’s papers with his advisor, “Distributionally risk‐receptive and risk‐averse network interdiction problems with general ambiguity set,” was published in the international journal, Networks, and he presented the research at the INFORMS Annual Meeting 2022.

At the INFORMS Annual Meeting in 2023, Kang presented another of their papers, “Distributionally Ambiguous Multistage Stochastic Integer and Disjunctive Programs: Applications to Sequential Two-player Interdiction Games,” in the student poster competition and in a poster session.

Projected to graduate in 2025, he will be exploring various opportunities to continue his research.