We are interested in the development of new algorithms at the intersection of optimization and machine learning. A particular focus is on integrating discrete aspects (as decisions can be discrete and continuous), which have traditionally been ignored as they often lead to very hard (almost always NP-hard) non-convex optimization problems. However, in recent years solvers for such discrete optimization problems (e.g., SCIP) got very fast and at the same time non-convex optimization problems have become more important due to their additional expressive power.