One exciting application domain for machine learning is the healthcare domain. In this domain, we have worked on designing RNA and decoding brain signals from electroencephalography (EEG). For RNA design, we used automated reinforcement learning to reach massive speed-ups and better performance compared to other methods in the field. For brain-signal decoding from EEG, we pioneered the use of modern deep learning, reaching at least as good decoding accuracy as well-established feature-based methods on movement-related decoding and automated diagnosis, while also reaching a deeper understanding of the information the deep networks extract from the EEG signal.

Find more information on these pages: