University of Hamburg is contributing to research area C with two projects. First, we investigate improved algorithms that combine tracking and calorimeter measurements from LHC detectors with timing information for improved precision, especially in the presence of a large number of simultaneous proton-proton collisions (pile-up). For this we develop neural network architectures that scale well and combine data from different sources. Secondly, we work to gain a clearer understanding of the decision process of deep neural networks in a physics context. Here we combine insight from physical variables with the latent space representation of data inside neural networks. Beyond these topics, the groups at UHH also work on other application of machine learning to particle physics, such as the development of better high-level tagging algorithms as well as model independent searches for new physics.