A0817
Title: Experimental design and data-driven optimization in complex systems
Authors: Matteo Borrotti - University of Milan-Bicocca (Italy) [presenting]
Abstract: Design of experiments (DoE) has long been a cornerstone in planning efficient data collection and enhancing model-based decision-making. However, modern complex systems, characterized by high dimensionality, costly evaluations, and multiple competing objectives, require extending the classical DoE toolbox with new computational and data-driven approaches. The foundations of DoE are revisited, and it is explored how emerging tools, such as Pareto-based selection, advanced optimization criteria, and algorithmic search methods, can be integrated into experimental design strategies. Through illustrative examples and recent developments, it is shown how these methods can support the design of experiments in challenging scenarios, from industrial processes to simulation-based optimization. Finally, open challenges and promising directions where data-driven optimization can enhance the role of DoE in the analysis of complex systems are discussed.
