RG Computational Psychiatry

This research group addresses current questions and challenges in psychiatric research with modern computational techniques from artificial intelligence, machine learning, and statistics. One focus lies on the detailed description of functional alterations in psychiatric disorders with statistical generative models such as, for instance, the characterization of behavioral learning deficits with reinforcement learning models, or the assessment of altered neuronal dynamics during experimentally induced cognitive processes with state space models and recurrent neural networks. On the other hand, we are exploring the capacity of machine learning techniques to predict functional alterations and disease trajectories based on big (multimodal) data sets, including neuroimaging and genetic data, or (longitudinally) in ambulatory assessments. Moreover, the research group represents a hub between experimental and theoretical neuroscience, with the aim of applying and advancing methods developed in our department specifically for psychiatrically relevant questions, and thereby promoting the development of new and intelligent psychotherapeutic interventions.