Decision-making methods under Uncertainty in Discrete Multi-objective Optimization

This dissertation investigates decision-making methods in Uncertain Discrete Multi-objective Optimization Problems (UDMOPs), where uncertainty arises in both objective function and constraint coefficients. The study pursues three main goals: (1) constructing sensitivity regions in the objective space to handle objective-wise uncertainty, (2) constructing sensitivity regions in the decision space to handle feasibility uncertainties, and (3) developing methods to sort, group, and prune uncertain solutions based on their similarity. Each goal proposes a method to explore uncertain solutions and quantify their level of uncertainty. Based on this, solutions are classified as low and high-risk solutions, according to the Decision-Maker (DM)'s preferences and risk tolerance. The proposed approaches employ stochastic optimization techniques to identify low and high-risk solutions, enabling risk-averse decision-making. Numerical experiments, including a real-world application, and benchmark comparisons, show that low or high-risk solutions under uncertainty can outperform the efficient solutions from deterministic model. Overall, the methods provide a more consistent and informative decision support system for DMs under uncertainty.

Grantor dc:publisher

University of Tennessee at Chattanooga

Year dc:date.available

2026

Author dc:creator

• Aththanayake, Chathuri Malee

Contributors dc:contributor

• Weerasena, Lakmali
• Ebiefung, Aniekan; Bandara, Damitha; Ma, Ziwei
• College of Engineering and Computer Science

• Decision making--Mathematical models
• Sensitivity theory (Mathematics)
• Stochastic programming
• Uncertainty--Mathematical models

Licence

• http://rightsstatements.org/vocab/InC/1.0/

Language dc:language

English, eng

Repository record dc:identifier

https://scholar.utc.edu/theses/1016

OAI identifier oai:identifier

oai:scholar.utc.edu:theses-2204