Conducted in collaboration with the Turkish Society of Nephrology's Renal Disaster Relief Task Force, the research focuses on improving patient access to life-sustaining treatment under emergency conditions.

My primary contributions focused on the development of a prototype decision support system that translates optimization models into a practical tool for disaster coordinators, and on the design of the Iterative Constructive Heuristic (ICH), a scalable solution algorithm that rapidly produces high-quality treatment plans for large real-world instances.

The project introduces a novel regression clustering framework that simultaneously identifies customer price sensitivities and estimates pricing models directly from historical quotation data.

The core contribution of the research is the development of the Iterative Refit Classifier (IRC), a heuristic regression clustering algorithm designed to solve large-scale pricing problems efficiently while maintaining high predictive performance.

To further improve accuracy without sacrificing interpretability, I also proposed a teacher-student framework that distills complex machine learning models into an explainable pricing methodology. Together, these methods enable pricing analysts to identify heterogeneous price elasticities across quotation groups and support more informed, data-driven pricing decisions in industrial B2B environments.