Despite the outstanding success of deep learning-based methods, most of the research on deep neural networks is empirically driven and mathematical foundations are largely missing. Moreover, in several special but important cases these techniques dramatically fail under small perturbations such as adversarial examples in image classification, which calls for improvements driven by a theoretical underpinning.
Successful proposals address a genuine contribution to the understanding and the theoretical foundations of deep learning along the following three complementary points of view:
- the statistical point of view regarding neural network training as a statistical learning problem and studying expressivity, learning, optimisation, and generalisation,
- the applications point of view focusing on safety, robustness, interpretability, and fairness, and
- the mathematical methodologies point of view developing and theoretically analysing novel deep learning-based approaches to solve inverse problems and partial differential equations.
Proposals must be written in English and submitted to the DFG by 30 November 2020 via "elan" the DFG’s electronic proposal processing system.