Machine learning
Given the vast amounts of incoming data from astrophysical surveys, machine learning is becoming a powerful tool in the search for dark matter. We have explored a variety of applications of machine learning to data from the Gaia satellite, which is mapping the positions and velocities of an unprecedented number of stars in the Milky Way. In one application, we trained a neural network to identify stars that were dragged into the Milky Way from other, smaller galaxies. By analyzing the resulting catalog of ‘‘accreted’’ stars, we discovered the Nyx stream—-a vast prograde stellar stream near the Sun. In a separate application, we trained a neural network to guess a star’s missing radial velocity based on its other measured parameters. This technique enables one to vastly extend the number of stars in the Gaia catalog with complete phase-space information.