The CAIDAS-Chair of Computer Science - Machine Learning for Complex Networks adresses new data science and machine learning techniques for complex systems that can be modelled as graphs or networks. We further use network science techniques to study open questions in biology, empirical software engineering, and computational social science. Our approach is quantitative, data-driven and interdisciplinary, combining methods from network science, machine learning, mathematics and physics.
Apart from (statistical) learning techniques for graph-structured data, a current focus of our chair is the use of higher-order graph models to better understand causal structures in time series data on complex systems, with applications in biology, ecology, information systems, and social sciences. This novel direction of research in network science has major implications for our understanding of complex systems, both in terms of theoretical foundations as well as in terms of machine learning methods. A summary of our approach to tackle this issue has been published in Nature Physics.
Our Chair has an international and interdisciplinary focus. Apart from developing new methods and applications of machine learning in relational data, we address issues that are fundamental for our understanding of complex systems across disciplines. Our research results have been published in leading physics journals like Physical Review Letters, Phys Rev E, or Nature Physics, in top-tier data science and machine learning venues like SIGKDD, NeurIPS, The Web Conference or Learning on Graphs (LoG), as well as in prestigious software engineering venues like ICSE, MSR, or Empirical Software Engineering.
If you are interested to work with us, please have a look at current openings.
