Scientists at the Manoa Institute of Astronomy (IfA) at the University of Hawaii have used AI to create the world’s largest 3D catalog of stars, galaxies and quasars.
The team developed the map using an optical survey of three-quarters of the sky created by the Pan-STARRS Observatory in Haleakala, Maui.
They train an algorithm for identifying celestial objects in the study by providing it with spectroscopic measurements that provide a final classification of objects and distances.
“Using a state-of-the-art optimization algorithm, we used the spectroscopic set of nearly 4 million light sources to teach the neural network to predict source types and distances of galaxies, while correcting the disappearance of light from dust in the Milky Way.”
This allowed the neural network to achieve a classification accuracy of 98.1% for galaxies, 97.8% for stars and 96.6% for quasars.
[Read: NASA releases images of the first craters on Mars discovered by AI]
The team says their catalog is twice the size of the previous largest map of the universe, which was created by the Sloan Digital Sky Survey (SDSS) and covers a third of the sky.
A preliminary, smaller version of the catalog has already helped find the largest cavity in the universe – a possible cause of the Cold Spot, a vast and mysteriously cold area of the sky.
The entire 300 GB database can now be downloaded as a computer-readable table.
Published on October 12, 2020 – 11:52 UTC