The recent boom in machine learning and artificial intelligence brought countless technological advances, but also many mathematical challenges. On the one hand, it is necessary for the theory about practical methods to catch up with their application, what passes through a better mathematical understanding of the methods. On the other hand, in order for continuing developing technologies, new mathematical theories need to be established as a basis for innovative methods. These mathematical needs make topics in mathematical data science cutting-edge research topics.

Projects will involve mathematically analysing data science methods from the point of view of statistical learning theory, approximation theory and optimisation. They might focus exclusively on mathematical abstractions that might be applied to data science methods, or also consider the development of new methods based on the theory.

References

Marcondes, D.; Simonis, A.; Barrera, J. Back to basics to open the black box. Nature Machine Intelligence. 2024