Roland Szakacs

•Performing scientific research at Europe's foremost astronomical observatory and supporting international collaborations. • Transformation, quantitative / statistical analysis and physical modelling of: Noisy (3D) spectroscopic data and Magnetohydrodynamical numerical simulations. • Development of deep learning models for feature detection and localization in large noisy 1D spectroscopic datasets. • Authored 9 publications (accepted / submitted) in peer-reviewed journals (3 first author, 6 co-author).

• Development of numerical chemistry models for simulations of galaxy formation and evolution. Included a detailed study on how specific model / parameter modifications impact the science result. • Creating large datasets based on inputs and outputs of numerical chemistry models and applying machine learning models to estimate these outputs.

• Scientific analysis of complex, 3D-spectroscopic astronomical observations and magnethydrodynamical simulations. • Physical interpretation of observations and simulations in the context of galaxy formation and evolution. • Highlight: Development of deep learning models for the analysis of noisy 1D spectroscopic data.

• Gained substantial analytical, quantitative, and problem-solving skills via various courses in Astrophysics, Statistics and Machine learning. • Highlight: Exploring deep learning models to approximate results of numerical simulations of astrochemical phenomena.

• Gained substantial analytical, quantitative, and problem-solving skills via various courses in experimental and theoretical Physics, Mathematics, and hands-on laboratory work. • Highlight: Building a pipeline which automatically scrapes data from websites and performs physical parameter estimations.