Machine Learning Meets Astrophysics

LLNL Data Science Institute Seminar Series

The LLNL DSI sponsored a seminar on May 22, 2018, featuring Dr. Andreas Zoglauer of the UC Berkeley Institute for Data Science. Zoglauer works with Berkeley’s Space Sciences Laboratory on the NASA-sponsored project COSI (the Compton Spectrometer and Imager), a balloon-borne gamma-ray telescope. COSI’s science objectives focus on galactic nucleosynthesis and the polarization of gamma-ray bursts caused by astronomical events such as neutron star mergers and core-collapse supernovae of heavily rotating massive stars. COSI’s 2016 flight around the southern hemisphere generated data that Zoglauer’s team continues to analyze.

According to Zoglauer, gamma-ray astronomy research relies heavily on data science and statistics. To analyze the data from COSI’s detectors, he developed an open-source toolkit called MEGAlib (Medium-Energy Gamma-ray Astronomy Library), which has applications beyond astrophysics in nuclear medicine and nuclear monitoring. MEGAlib enables researchers to perform Monte Carlo simulations of their detectors, reconstruct Compton events, and create images based on Compton scattering data. Zoglauer stated that COSI’s biggest computational challenge is generating up to 9-dimensional response files with Monte Carlo simulations for the reconstruction of all-sky images. Those simulations were performed on Berkeley Lab’s cori supercomputer.

With the help of data science undergraduates, Zoglauer is applying machine learning to COSI data such as random forests and neural networks. Research projects include determining photon paths in the germanium detectors, finding interaction locations in the detectors, and identifying not-contained gamma rays. Zoglauer outlined several lessons learned through his team’s work with machine learning tools, such as the importance of preparing data, splitting a big research question into smaller questions, and verifying that the trained neural networks have no “blind spots.” Researchers using machine learning algorithms should also expect “a lot of trial and error” in finding the best input data representation.

COSI is preparing for another flight in 2019–2020. An upgraded version, COSI-X, is planned for launch in 2022 with additional detectors, better shielding, and improved resolution.

Speaker(s)

Andreas Zoglauer

BIDS Alum - Data Science Fellow
Andreas Zoglauer is an astrophysicist working at the intersection between astrophysics & data science. As a BIDS-LLNL Data Science Fellow at UC Berkeley, he worked on the development of new data analysis tools and their application to hard X-ray and gamma-ray telescopes. His main project was COSI, a balloon-borne gamma-ray telescope which was built from ground up at Berkeley’s Space Sciences Laboratory. COSI had a record setting 46-day stratospheric balloon flight in 33 km altitude and observed gamma rays from Galactic positron annihilation and Galactic nucleosynthesis, as well as from pulsars, binaries, and black holes. Andreas leads the development of the universal open source calibration, simulation, and data analysis toolkit MEGAlib for X- and gamma-ray telescopes. MEGAlib can be applied not only to space telescopes but also to detector system on ground such as for nuclear medicine (e.g. hadron therapy monitoring) or environmental monitoring (e.g. the ARES & HEMI detector systems developed at LBNL). Within the MEGAlib framework, Andreas focused on applying the latest deep learning approaches to the data analysis of gamma-ray detectors and improving the image quality of Compton telescopes using high-dimensional responses created on Berkeley’s Cori supercomputer.  He received his PhD from the Technische Universität München, Germany, for the development of novel simulation and data analysis tools for the Compton-scattering and pair-creation telescope MEGA.