Research idea behind AstroAI-Lab

With the advent of (deep) neural networks, computers nowadays excel at tasks such as image or speech recognition, previously unthought to be solved by machines. At the same time, deep learning is becoming increasingly important for industry, engineering, natural sciences but also society. Therefore, security and equity concerns but also external constraints such as natural laws represent fundamental obstacles for the general breakthrough of conventional machine learning (ML). These shortcomings have triggered the development of research areas such as trustworthy AI or interpretable and explainable ML. Exploring and contributing to the scientific usage of these innovative developments is the subject of AstroAI's research.

Both, ML and computer simulations, share the goal of predicting the behaviour of a complex system using data analysis techniques and mathematical modelling approaches. Thereby, Astrophysical phenomena, such as galaxy formation, are inherently an interdisciplinary, massively multi-scale, multi-physics problem, commonly addressed with numerical models requiring high-performance computing facilities and millions of CPU hours. Nevertheless, scientific knowledge gain is limited by the amount of computing resources required to calculate all the relevant physics. Thus, there is a pressing need for a paradigm shift in the way we build and employ our numerical models. Thus, the scientific rational here is to explore how modern ML techniques can be incorporated into the scientific reasoning process in general and to obtain new insights into the physical processes of the formation and evolution of our Milky Way galaxy in particular. In order to fully exploit those innovative methods in the natural sciences we need to develop ML methods that are inherently interpretable and respect the laws of physics. Especially the field of theoretical/computational astrophysics has the potential to greatly benefit from these developments, while at the same time, driving advances in the field of ML.

The research of the AstroAI-Lab will ultimately combine the flexibility of ML tools with the predictability of classical numerical simulations to develop innovative and interpretable hybrid ML/simulation resources and algorithms which will certainly have societal and industrial impact well beyond the scientific scope of astrophysics. Evaluating their scientific impact and value, and releasing them as production ready tools to the scientific community is the agenda of AstroAI-Lab.