Robust Optimization for Integrated Healthcare Scheduling

Multi-resource MILP with budgeted uncertainty sets to jointly optimize surgical case planning and nurse-to-patient assignment — quantifying the Price of Robustness against surgical duration variability and nursing acuity spikes.

Read more about the project →

MILP Gurobi Python Robust Optimization Healthcare Operations

The Integrated Healthcare Timetabling Problem — A Systematic Review

Structured literature review of 50+ papers across Patient Admission Scheduling, Surgical Case Planning, and Nurse-to-Room Assignment — mapping the field from foundational metaheuristics to modern robust and stochastic models, and establishing the methodological frontier for fully integrated planning.

Read more about the project →

Literature Review Healthcare Scheduling Metaheuristics Robust Optimization IHTP

MSc Thesis — Multi-objective Robust Optimization for Remanufacturing SystemsWorking Paper

Multi-period, multi-objective robust MILP for hybrid manufacturing and remanufacturing planning under simultaneous uncertainty in return quality, processing times, and demand — solved with NSGA-II and validated on a real industrial case.

Read more about the project →

Robust Optimization Multi-objective Remanufacturing NSGA-II MILP

Tire Closed-Loop Supply Chain Network Design under UncertaintyWorking Paper

Working paper. Multi-objective MILP for designing a sustainable tire supply chain with two rounds of remanufacturing (retreading) — minimizing cost and environmental impact under stochastic demand and fuzzy return and production rates.

Read more about the project →

MILP Robust Optimization Stochastic Programming Remanufacturing

Large-Scale Optimization of Healthcare Logistics via Benders Decomposition

Benders decomposition algorithm implemented in GAMS to solve a blood supply chain network design problem — exploiting the natural strategic–operational decomposition to achieve tractability on instances where monolithic MILP fails.

Read more about the project →

Benders Decomposition GAMS Blood Supply Chain Healthcare Logistics MILP

Discrete-Event Simulation for Transportation System Performance Analysis

Arena-based discrete-event simulation of a multi-stage transportation system — generating utilization profiles, queue-length distributions, and throughput sensitivity curves to identify binding bottlenecks and evaluate the system-wide impact of operational interventions.

Read more about the project →

Arena Discrete-Event Simulation Queueing Theory Bottleneck Analysis Transportation

Machine Learning

Next-Basket Prediction using Bidirectional Transformers

Progressive enhancement framework for e-commerce next-basket recommendation — replacing a GRU baseline with a bidirectional BERT4Rec Transformer and structured data augmentation, achieving a 4× improvement in Recall@10 on the Retail Rocket dataset.

Read more about the project →

Transformers BERT4Rec Deep Learning Recommender Systems E-Commerce

Blockchain-Based Transparent and Sustainable Supply Chain Network Design

Analysis of a blockchain-enabled closed-loop supply chain framework — integrating IoT/RFID traceability, smart contract governance, and a multi-objective fuzzy optimization model to simultaneously balance economic, environmental, and social objectives.

Read more about the project →

Blockchain Supply Chain Sustainability RFID Smart Contracts Network Design

AI Use in Healthcare: Ethical Framework for Diagnostic Applications

Evaluation of ethical risks in AI-based medical diagnostic systems and design of a responsible framework built on contestability, explainability, accountability, privacy, fairness, and answerability — with a focus on contestability as the mechanism enabling meaningful human oversight.

Read more about the project →

AI Ethics Healthcare AI Explainability Contestability Deep Learning GRAD-CAM SHAP

Customer Lifetime Value Assessment via Weighted K-Means Clustering

Combined RFM scoring with a CLV-weighted K-Means algorithm to segment 2,239 grocery store customers — demonstrating that explicit lifetime value weighting produces sharper, more actionable segments than standard RFM classification alone.

Read more about the project →

K-Means RFM CLV Clustering CRM

Empirical Research

Structural Equation Modeling for Supply Chain Strategy and Environmental Performance

Two-part quantitative research project applying EFA, moderation/mediation analysis, and PLS-SEM to examine how purchasing recognition, long-term supplier commitment, and proactive environmental strategies drive supply chain performance across Turkish manufacturing firms.

Read more about the project →

PLS-SEM SEM SmartPLS Supply Chain Environmental Performance Quantitative Methods

Economic Incentives and Household Waste Reduction — A Prospect Theory Experiment

Controlled experiment testing fines versus subsidies on household waste generation over 10 days — confirming loss aversion: the fine (loss frame) produced a 28% reduction versus 8% for an equivalent subsidy, consistent with Prospect Theory predictions.

Read more about the project →

Prospect Theory Behavioral Economics Experimental Design Sustainability Waste Management

CRM Systems as Platforms for Democratizing Corporate Science

Philosophy of science paper arguing that Customer Relationship Management systems can be redesigned as epistemic interfaces — linking diverse stakeholder knowledge with corporate R&D agendas through frameworks of well-ordered science, hybrid forums, and epistemic justice.

Read more about the project →

Philosophy of Science CRM Epistemic Justice Science Governance Participatory Research