Abstract: 

Space Science with Python is an online tutorial series that combines Python, Astronomy, Data Science and Machine Learning in one learning curriculum. In recent years space missions like the James Webb Telescope (JWST), the Mars rover Curiosity, as well as private initiatives like SpaceX have sparked and enlarged the public attention and interest in space and our cosmic vicinity. Whether by images from the JWST or the Hubble telescope: impressive images are well received by amateurs and professionals alike.

Especially the Python community has proven in the last couple of years that dedicated developers create amazing contributions to complex domains like Deep Learning or Data Science. However, in the current Open Source world it appears that space and astronomy related data scientific topics and projects are less “approachable” than the topics addressed before.

In this tutorial we will dive into a particular space science / engineering domain: the calibration of space instruments. For this we take a dedicated look at calibration data from the so-called Cosmic Dust Analyzer (CDA) that was part of NASA's Cassini mission in the Saturnian system. Together, we will see how the data has been generated, explore their features and limits and will determine how deep learning can help us to create new state-of-the art calibration solutions for space missions.

Background Knowledge:

Docker, VS Code, Jupyter Notebook, Common Data Science Libraries like NumPy, Matplotlib, Pandas, ...

Bio: 

Thomas is a Senior Machine Learning engineer, working in the automotive industry since 2019. Before joining the Research & Development department of a large manufacturer he was conducting research activities in space science. In parallel to his studies in Astro- and Geo-Physics and later PhD program, he participated in 2 major missions: ESA's comet mission Rosetta/Philae and NASA's & ESA's Saturn spacecraft Cassini/Huygens; always with a special focus on cosmic dust. Additionally, he applies Machine Learning algorithms to analyse astronomy- and space-related data to derive new scientific insights or to create new methods for calibrating instruments. Besides his industry work, Thomas is a guest scientist at the Free University of Berlin, where he continues working on the Cassini-related datasets using Deep Learning. On his active YouTube channel Astroniz he shares his Python + Space Science + Machine Learning knowledge with a small community.