The core objective of this RTG is to bridge the gap between control theory (CT) and machine learning (ML) and to train a new generation of scientists with skills on both sides. Such scientists are urgently needed to face the challenges of modern artificial intelligence (AI) research and to establish datadriven discovery as the “fourth paradigm” of scientific discovery.
A Research Training Group is exactly the right format to realize this vision: It makes it possible to foster research in a modern field while at the same time ensuring sustainability by training scientists who can consolidate this new field over the course of their academic careers. At present, there is an apparent lack of such scientists. Engineering students typically have a strong background in calculus, differential equations, and scientific computing but less in optimization, data analysis, and ML. On the other hand, computer scientists and statisticians are well trained in these latter topics and skilled in programming but have little exposure to dynamical systems and control.
The ambition of METEOR is to train a new generation of researchers at the intersection of machine learning and control theory for complex dynamical systems, advancing research through exploring synergies, complementarities, and mutual benefits of both fields.