Optimization

The Optimization pillar contributes to data science through the development of advanced methods and frameworks for data-driven decision making and decision support. Our core competences include mathematical modeling, the design of heuristics and metaheuristics, machine learning–based hyperheuristic optimization frameworks, and multi-objective optimization. We focus on solving combinatorial and operations research problems arising in domains such as energy, transportation, and logistics. These problems often involve complex trade-offs, large-scale data, and strict operational constraints, requiring customized solution approaches that balance computational efficiency with decision quality. Our research bridges prescriptive and predictive analytics, enabling systems that learn from data while producing robust, actionable decisions. The Optimization pillar collaborates actively with other pillars to develop scalable optimization tools, integrate data-driven models into decision-making processes, and contribute to comprehensive, end-to-end data science solutions.