Christina Kreisch: using machine learning tools to probe the universe’s evolution

Written by
Sharon Adarlo
Aug. 17, 2021

Christina Kreisch, 28, doctoral student


Kreisch is a doctoral student in astrophysics and is slated to defend her thesis in August. She holds a Charlotte Elizabeth Procter Fellowship, awarded to exceptional graduate students in their later years of study. In addition, she is completing the graduate certificate program at the Center for Statistics and Machine Learning (CSML) and another one on computational and information science from the Princeton Institute for Computational Science and Engineering. She is also pursuing a teaching transcript at the McGraw Center for Teaching and Learning.

Before starting her doctoral studies, she earned a master’s degree in astrophysics from Princeton in 2018. In 2015, she earned a bachelor’s degree in physics and mathematics at Washington University in St. Louis.


Kreisch's research interests lie in cosmology, a branch of astronomy that concerns itself with the universe’s origin and its evolution. She marries this interest with machine learning, which scientists have increasingly come to rely in recent years in order to interpret cosmological data.

I work at the intersection of theory and computation and cosmology,” said Kreisch.

Her thesis is made up of three parts. The first part is theory and computationally-driven and concerns itself with neutrinos and their interactions in the early universe. This research, “The Neutrino Puzzle: Anomalies, Interactions, and Cosmological Tensions,” appeared in the journal Physical Review D in April 2019 and garnered 170 citations. Kreisch is first author.

“We essentially worked on a new theory relating to particle physics, specifically neutrinos. And we said, ‘let's see what happens if we let them interact in the early universe and how does that impact the evolution of the universe?’” she said.

Kreisch tested this theory using cosmological observations and Markov Chain Monte Carlo based Bayesian data analysis and found that her results had surprising insights, especially in possibly helping solve controversies such as the Hubble constant, the numerical rate in which the universe is expanding. There are two different methods at arriving at this answer with different results, a problem that has bedeviled scientists.

The second part of her thesis concerns itself with cosmic voids, which are vast, almost empty areas between superstructures of gravitationally-bound galaxy superclusters. (A paper on this just came out with Kreisch as first author. Read here.) The third part of her thesis relates to machine learning, touching upon topics such as using graph neural networks on cosmological data.

The CSML certificate has been useful for Kreisch because it has enhanced her research skillset, she said.

“Machine learning is becoming a very important tool in astrophysics, particularly since the field is so data rich and we have instruments coming online that will be collecting terabytes of data per night. So, it's really imperative to build this skill set in order to extract as much information as we can from this rich amount of data we're fortunate to have in our field,” she said.

Besides her time at Princeton, Kreisch’s experience includes visiting researcher at the Observatories of The Carnegie Institution for Science, research scholar at the Max Planck Institute for Astrophysics, research fellow at NASA’s Jet Propulsion Laboratory, and research intern at the Harvard-Smithsonian Center for Astrophysics.

Extracurricular Activities:

Kreisch has been volunteering as a ReMatch and USRP mentor to women interested in astrophysics and cosmology on campus. ReMatch is a Princeton research-mentoring program designed to connect undergraduate students, graduate students and postdoctoral researchers.

For Fun:

Kreisch enjoys riding horses, singing and gardening.