(ORDO NEWS) — DR3 “Gaia” was recently released, the latest dataset containing over 1.8 billion objects.
This is a huge amount of data that needs to be processed, and one of the most efficient ways to do this is through machine learning.
A group of researchers used a supervised learning algorithm to classify a specific type of feature found in the dataset.
The result is one of the world’s most comprehensive catalogs of astronomical objects known as variables.
“Gaia”, which has been observing vast areas of the sky for a long time, is especially adept at detecting them.
She discovered about 12.4 million variable sources, about 9 million of which were stars. More than 3 million were either active galactic nuclei or galaxies.
12.4 million out of 1.8 billion is only about 0.6% of the total number of observed objects in DR3.
However, they may contain information that will help astronomers understand the causes of certain types of variability.
To sort 12.4 million objects, the researchers turned to machine learning. In particular, they used a controlled classification method.
Essentially, this means that they had a person help an artificial intelligence algorithm identify the features of a certain classification and then provide feedback on whether the object met the criteria for being classified in that category.
The resulting data set is the world’s most complete catalog of variable astronomical objects and tools for their scientific processing.
These kinds of data releases are precisely the milestones that are helping astronomy move forward. The Gaia Telescope still has a lot to do, with DR4 coming sometime after 2025.
That way, astronomers will have plenty of time to study all of the DR3 data in detail before the next mass data release.
Contact us: [email protected]