Dissertation Defence: Intelligent Decision Support Systems Design for Real-Time Management of Large-Scale Optimization Problems
November 12 at 9:00 am - 1:00 pm
Mahsa Mohammadi, supervised by Dr. Babak Tosarkani, will defend their dissertation titled “Intelligent Decision Support Systems Design for Real-Time Management of Large-Scale Optimization Problems” in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Mechanical Engineering.
An abstract for Mahsa Mohammadi’s dissertation is included below.
Examinations are open to all members of the campus community as well as the general public. Please email babak.tosarkani@ubc.ca to receive the Zoom link for this defence.
ABSTRACT
This thesis is focused on developing intelligent decision support systems (IDSS) to tackle large-
scale optimization problems in various domains including e-commerce businesses (last-mile delivery),
global supply chain, and healthcare (vaccine distribution, mask distribution and recycling). Ad-
vanced mathematical modeling, data-driven techniques, and innovative policies are employed to
improve real-time decision-making under dynamic nature of the world. In Chapter 2, the last-mile
delivery problem is addressed by developing a multi-stage stochastic dynamic risk-averse model.
Three innovative policies including mobile depots, crowd-shippers, and hyper-local drivers are in-
tegrated into this model. Crowd-shipper availability is estimated using a deep learning model,
and an online reinforcement-driven adaptive optimization method is proposed. Improved deliv-
ery times and reduced emissions are demonstrated through validation using interactive multi-agent
simulation. In Chapter 3, a multi-stage stochastic dynamic model incorporating hedging policies
is presented to enhance global supply chain resiliency. A discount rate model for optimal supplier
financing is introduced, and an accelerated parallel stochastic dual dynamic integer programming
(PSDDiP) algorithm is proposed, achieving faster computation and superior performance in large-
scale scenarios. Chapter 4 involves the development of a data-driven decision support system for
vaccine distribution, focusing on the dynamic allocation of mobile and stationary public health units
(PHUs). The augmented data-driven robust optimization (ADDRO) approach is introduced to
handle limited data availability, thereby improving performance. Additionally, a dynamic PHU
allocation algorithm is developed to ensure equitable access to vaccination, optimizing distribution
costs and coverage. In Chapter 5, the design of a sustainable closed-loop supply chain network for
mask distribution and recycling during the COVID-19 pandemic is focused upon. A robust multi-
objective mixed-integer linear programming model is developed to handle uncertain demand and
processing times. By applying a sample average approximation (SAA) and scenario decomposition
(SD) methodologies and proposing a graph theory-based clustering (GTC) algorithm, computa-
tional efficiency is improved, and sustainability aspects are integrated into the personal protective
equipment (PPE) supply chain design. In summary, these chapters advance IDSS s by addressing complex and large-scale problems under dynamic nature of the real-world. Sustainability, resiliency,
and equitable access are emphasized, enhancing the practical relevance and societal impact of the
research.
