Welcome. Tervetuloa!
Michael Emmerich
Professor of Intelligent Computing · PS Fellow
University of Jyväskylä, Finland
Guest Scientist, Leiden University, The Netherlands
I study multiobjective optimization, decision analysis and intelligent computing. My work combines mathematical foundations with methods that help people understand trade-offs and make better decisions.

Research
- Multiobjective optimization and decision analysis. Methods for exploring conflicting objectives, preferences and complex decision spaces.
- Machine learning and AI. Gaussian process regression, optimal model ensembles and AlphaZero-style deep reinforcement learning.
- Geometry and diversity. Metric geometry, mathematically motivated diversity measures and connections to computational geometry.
Applications of my research include chemical technology, engineering design, medicinal chemistry, epidemiology and network science, sustainable forest management, and operations research.
I support the Leiden Declaration on Artificial Intelligence and Mathematics.
Publications & learning
Search the publication database for summaries, DOI links, BibTeX citations and available PDFs. You can also find my publications on Google Scholar.

Multicriteria Optimization and Decision Making
My free introductory reader on multicriteria optimization and decision making.
Read the open-access book (PDF)
Many-Criteria Optimization and Decision Analysis
Our 2023 Springer book explores results and challenges in optimization with many objectives. Edited by D. Brockhoff, M. Emmerich, B. Naujoks and R. Purshouse.
What makes computing intelligent? Read my short AI tutorial (PDF).
Multicriteria healthcare
Research on applying multicriteria methods to healthcare, with Atta and Basto-Fernandes.

Data, support materials & software
Explore research code, datasets, benchmarks and supporting documents, including materials from the former MODA group at Leiden University. The catalogue provides searchable descriptions, source credits, links to related publications and available ZIP downloads.
Talks & past events
Shaping Expected Hypervolume and R2 Improvement
25 May 2026 · MCDM, Wuppertal, Germany
Shaping Expected Hypervolume and R2 Improvement for Preference-Guided Bayesian Optimization: Exact Integration, Pareto Compatibility, and Variance Monotonicity. International Conference on Multiple Criteria Decision Making, Bergische Universität Wuppertal.
Download the presentation slides (PDF)
Modern Machine Learning Technologies (MoMLeT) — past workshop series, which ended in 2025.
Beyond research
Alongside my professional work, I enjoy nature photography, painting, recreational mathematics and programming. I live with my family in Jyväskylä, Central Finland.
About & connections
I previously served as Lead AI Scientist at silo.ai in Finland and Associate Professor at Leiden University. I have also conducted research at institutes and universities in the United States, the Netherlands, Italy, Portugal and Germany.
My approach to responsible AI
I advocate the responsible use of AI. Following computer science pioneer Fred Brooks, I want to develop AI that amplifies human intelligence and helps people solve challenging problems. These include healthcare, a sustainable future for humans and our planet, and complex design problems in engineering and chemistry.
I see particular potential in multiple criteria optimization and analysis: it involves human decision makers and assists them in navigating complex decision spaces.
Research groups
- Multiobjective Optimization Group, University of Jyväskylä
- Current Leiden Natural Computing group
- Former MODA group, Leiden University — website archive; downloadable research materials are preserved in the materials catalogue.
Societies & associations
Contact
For professional enquiries, please use the contact details on my University of Jyväskylä profile.