"My research is located on the overlapping areas of control engineering and artificial intelligence. By leveraging the capabilities of modern deep learning methods, the application of complex control and decision-making algorithms is rendered possible even on hardware with limited computational resources. One major aspect of my work is to provide means of stochastical and optimization-based nature to provide guarantees on the safety and performance of neural networks controllers derived via deep learning. The techniques developed in my research are widely applicable. One exemplary usage is the embedded implementation of optimization-based sensor fusion problems as the ones explored in Felix Fiedler's research project."
Period of contribution: 2018 - 2019
"I am a PhD student in Theoretical Chemistry. I am familiar with all the theories and program architectures used by Kanishka Singh, as well as the tools for analysing and interpreting chemistry-related results (i.e. the spectra). Being myself a programmer for GPU and quantum compute systems and expert in many programming languages, I provide general support on Kanishka's project."
"In my research I develop and apply deep learning models to study gene regulation. I then incorporate the predictions of these models into genetic association tests, which allows me connect gene-regulatory mechanisms and disease. Specifically I am interested in kernel-based association tests that incorporate prior knowledge, and interpretable deep learning models that leverage data from multiple species."
"My research focus lies on forward modeling of the interior dynamics of rocky planets. I combine global-scale geodynamic models with observational constraints to improve our understanding of the interior evolution of Mars and Venus. Within the HEIBRiDS research program I closely collaborate with Siddhant Agarwal to combine machine-learning techniques with geodynamic modeling that will help to constrain key parameters for the thermal evolution of terrestrial planets."
"In my doctoral project, I deal with the subject areas of E-Mobility & Smart Grids and Railway Delay Management and apply methods of applied operations research and data science. I work in a related topic with Paolo Graniero, with whom I exchange ideas both technically and methodically."
"I am Associated HEIBRiDS Doctoral Researcher, employing Implicit Generative Models and Latent Variable models to learn meaningful representations of multi-omics datasets and to analyze them for a secondary biological application. My focus is on tailoring the solutions to cope with the limitations of biological datasets, interpretability of the results, as well as model performance and reusability."
"My research interests include data stream processing, sensor data analysis, and data acquisition from sensor nodes. I have authored several publications related to data stream gathering, processing and transmission in the Internet of Things. In the context of HEIBRiDS, I work with Sergey Redyuk in the area of scalable data analysis".
"My PhD project can be regarded as experimental counterpart of Peter Tillmann's PhD project in data science. Peter is simulating the energy yield of different kinds of solar modules using realistic weather data including forecasts by machnine learning methods. I develop light management textures for these solar modules, aiming at generating energy yield data from photovoltaics outdoor testing facilities. At the end these experimental energy yield data will serve as input for Peter's computations."