Michael Emmerich, a Finnish/German computational scientist, specializes in Applied Mathematics and Computer Science. He is Professor in Multiobjective Optimization at the Faculty of Information Technology, Jyväskylä University. Formerly a Lead AI Scientist at Silo.ai, and Associate Professor at Leiden University, Michael has also conducted research at IST Lisbon, TU Dortmund, RWTH Aachen, FOM/AMOLF Amsterdam, and has been a visiting professor at Lviv University, Princeton University, and the University of Jyvaskyla.
In 2005, Michael Emmerich completed his Doctorate in the Natural Sciences under Hans-Paul Schwefel and Peter Buchholz. His research focused on integrating machine learning with uncertainty quantification in multiobjective and constrained nonlinear design optimization, leading to the development of now widely-used methods such as SMS-EMOA, Large-Scale Gaussian Process Regression, and Expected Hypervolume Improvement.
He is also noted for his work on efficient algorithms in the context of Indicator-based and Bayesian Multiobjective Optimization, as well as on the introduction of the concept of set-indicator derivatives which are used, for instance, in the Hypervolume Newton-Raphson Method for Constrained Non-Linear Multiobjective Optimization.
Besides has contributed to applied machine learning, optimization, and applied operational in areas like drug discovery, computational biology, sustainable forestry, building design, aviation logistics, and chemical technology. Michael has supervised over seventeen Ph.D. students who have gained prominence in academia. He has authored 70+ journal articles and over 100 papers in peer-reviewed conference proceedings, chaired international conferences like EMO 2023, EVOLVE 2023, and initiated the Modern Machine Learning Technologies and Data Science CEUR-WS Series (2018-2023). With an h-index of around 40 (44 on Google Scholar), he has edited over ten books in multiobjective and global optimization.
Michael's current research interests encompass several critical areas:
1. Exploring the fundamentals of Multicriteria Optimization Algorithms, with a specific focus on Computational Geometry Algorithms in Indicator-Based Multicriteria Optimization and Bayesian Multiobjective Optimization.
2. Developing Application-Domain Specific Multicriteria Optimization Algorithms for applications such as Spatial Design, Chemical Engineering, Complex Networks, Sustainable Forestry, Logistics, Routing, and Truck Loading.
3. Investigating Complex Networks Dynamics and Modulation, particularly in the context of sensor placement and the contagion of epidemics based on Spectral Methods and Random Graph Theory.
4. Delving into Machine Learning, including Kriging, Gaussian Process models (including Cluster Kriging for big data), clustering, deep reinforcement learning, and optimally weighted model mixtures.
