(Machine) Learning the Genealogy of the Milky Way

Galaxies form and grow by merging with other galaxies, making the formation history of a galaxy resemble that of a family tree. Our galaxy, the Milky Way, is no exception. With recent telescopes like the Gaia space mission, we can finally build the Milky Way’s family tree.

In this Presidential Lecture, Lina Necib will discuss how we can use machine learning techniques to unveil the secrets of the merger history of our galaxy, including clustering techniques that group stars by their original galaxies, neural networks that separate stars with galactic and extragalactic origins, and anomaly detectors that uncover faint signatures of old mergers. Putting all this work together is a step towards building our galaxy’s family tree, she says.

Necib is originally from Tunisia. She received a B.A. in math and physics from Boston University in 2012 and a Ph.D. in particle physics from MIT in 2017. She was then a Sherman Fairchild Fellow at Caltech starting in 2017 before starting as an assistant professor of physics at MIT in July 2021. Her research interests in theoretical astrophysics center around dark matter.