Multi Objective Optimisation

Multi-objective optimization (MOO) tackles problems with multiple, often conflicting, objectives, aiming to find optimal trade-offs rather than a single "best" solution. Current research emphasizes efficient algorithms, including evolutionary algorithms (like NSGA-II and its variants), Bayesian optimization, and increasingly, machine learning-based surrogates to accelerate computationally expensive simulations. These advancements are crucial for addressing complex real-world challenges across diverse fields, from engineering design (e.g., optimizing wind farm layouts or draft tube designs) to machine learning (e.g., fairness-aware classifiers) and even personalized food production. The development of robust and efficient MOO methods continues to be a significant area of focus, driven by the need to solve increasingly complex optimization problems.