Rigour before hype
Mathematical modelling, sound experimental design and careful evaluation underpin our work in AI, optimisation and scientific computing.
Scientific Machine Learning →We connect mathematical modelling, data analysis and Artificial Intelligence to study complex phenomena, develop rigorous methods and translate research into measurable real-world value.
M.O.D.A.L.Artificial Intelligence becomes more meaningful when strong mathematical foundations meet reliable data, transparent experimentation and the questions that matter beyond the laboratory.
M.O.D.A.L. brings together expertise in numerical analysis, machine learning, scientific computing and distributed systems. Our goal is not simply to apply existing algorithms, but to advance methods, test their limits and build knowledge that others can reproduce and extend.
See the scientific output →Mathematical modelling, sound experimental design and careful evaluation underpin our work in AI, optimisation and scientific computing.
Scientific Machine Learning →We study privacy, robustness, unlearning and decentralised learning, with attention to what models retain, forget and generalise.
Federated Learning →Research grows through partnerships, doctoral training and exchange with academic, industrial and public-sector communities.
Our collaborations →An iterative process that balances methodological novelty with evidence and usefulness.
Formalise the problem, its assumptions and what success would actually mean.
Design mathematical and computational methods grounded in the state of the art.
Benchmark against relevant alternatives and study behaviour across realistic conditions.
Publish findings, support reproducibility and pursue responsible knowledge transfer.
People make the research possible.Mentorship, doctoral research, open exchange and international scientific dialogue are essential parts of M.O.D.A.L. We value autonomy, collaboration and the freedom to pursue difficult questions.