Skip to main content

CIERA researchers showcase data-driven discovery at CoDEx 2025

On April 8, 2025, Northwestern Information Technology (NUIT) hosted its annual CoDEx Symposium, a full-day event that brings together researchers from across campus to highlight data-driven and computational work. Open to faculty, postdocs, students, and staff, the symposium is designed to spark collaboration and showcase the innovative research happening throughout the Northwestern community. As in previous years, CIERA participants were actively involved, sharing their work through talks and posters. 

 

Keynote Speaker Adam Miller, assistant professor, Department of Physics and Astronomy, Weinberg College of Arts and Sciences, kicks off his presentation, “Ten Billion Galaxies! Ushering In the Petabyte Era in Astronomy and Astrophysics” in McCormick Auditorium. Image credit: Northwestern IT / CoDEx 2025 Recap

During the morning keynote session, CIERA Professor Adam Miller presented his talk “Ten Billion Galaxies! Ushering In the Petabyte Era in Astronomy and Astrophysics.” His presentation highlighted the revolution underway in wide-field, time-domain surveys—large observational campaigns that repeatedly image the night sky to capture transient astrophysical events. Miller emphasized the transformational potential of upcoming projects like the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST), which is expected to discover 2,000 new exploding stars every night.

 

Miller shared how his research group applies machine learning models—trained entirely on Northwestern’s Quest computing cluster—to analyze the massive data streams from these surveys. This work recently resulted in the first fully automated detection, classification, and public announcement of an exploding white dwarf. He also outlined challenges and successes in applying machine learning to real-time astronomical data, offering practical insights such as the importance of automating repeatable tasks while preserving essential roles for human judgment.

 

CIERA graduate students enjoy the keynote presentation, Ten Billion Galaxies! Ushering In the Petabyte Era in Astronomy and Astrophysics, given by Adam Miller. Image credit: Northwestern IT / CoDEx 2025 Recap

“Part of the reason I am deeply fascinated by astronomy is that the Universe is so large it is not ‘difficult’ to observe something that no other human has ever seen before. I first experienced this 20 years ago as an undergraduate working on a summer project at Northwestern, but this is still happening today, as Dr. Steve Schulze, a research associate working in my group, recently announced the discovery of an entirely new and completely unanticipated type of stellar explosion. ”
Adam Miller  Assistant Professor, Department of Physics and Astronomy Weinberg College of Arts and Sciences and member CIERA member

 

Tjitske Starkenburg

CIERA Prof. Tjitske Starkenburg presented her talk “Learning the Galaxies to Learn the Universe” during the morning parallel sessions. She discussed how astrophysicists use visiblecomponents of the universe, such as stars and galaxies, to explore fundamental questions about dark matter, dark energy, and cosmic origins. With large simulation suites, they compare possible universes to the observed large scale structure as traced by millions of galaxies. Starkenburg highlighted how galaxy formation itself presents major uncertainties that are often not feasible to include in large-scale simulations. Her talk presented a collaborative effort to incorporate uncertainties in both cosmological parameters and galaxy formation physics into combined statistical inference methods, illustrating how small-scale physics and large-scale structure connect to produce universes like the one we observe.

CIERA’s Graduate student Chang Liu presented a poster on his open-source software, HostSub_GP, a Bayesian toolkit designed to improve the precision of host galaxy subtraction insupernova spectroscopy. His method uses archival galaxy imaging and models background light as a two-dimensional Gaussian process, significantly reducing bias in spectroscopic measurements and outperforming traditional techniques.

Read the full article: Northwestern Research Achievement and Discovery Celebrated at CoDEx 2025