Galaxies and Cosmology Seminar

Dec

2025

12:00 pm – 1:00 pm

In Person (view details)

Learning the Galaxy-Halo Connection with Simulation-Based Inference

Description

Learning the Galaxy-Halo Connection with Simulation-Based Inference

Galaxies form and evolve within host dark matter halos. Although these halos cannot be observed directly, their properties—such as mass and formation history—are strongly connected to the observable characteristics of the galaxies at their centers. This connection is key to understanding the physics that shapes galaxies. In this talk, I will show how we can use machine learning to extract information about host halos encoded in observable features of galaxies and their environments. With this approach, we build models that predict halo masses, concentrations, spins, and accretion rates (with associated uncertainties) directly from observables. This framework provides a data-driven way to probe the galaxy–halo connection and test theories of galaxy formation in the universe.

Location

PMA 15.216B

Audience

Other Events in This Series

Mar

2026

Galaxies and Cosmology Seminar