Thanks to artificial intelligence and amateur astronomers, approximately 400,000 spiral galaxies and 30,000 ring galaxies have been discovered in archival data from the Subaru Telescope.
per week
Galaxies display a variety of shapes that reflect the history of individual galaxies. There are so many galaxies in data sets obtained from facilities like the Subaru Telescope that astronomers cannot classify them all by hand.
In a citizen science project called ‘Galaxy Cruise’, professional astronomers asked more than 10,000 volunteer citizen astronomers to make classifications. But a statement from the Subaru Telescope said the classification still took time, even after being divided among thousands of volunteers.
Artificial intelligence can perform classification faster, but it first needs to be trained on a list of human-prepared classification examples.
In this research, a team led by Rhythm Shimakawa, Associate Professor at Waseda University, trained the AI on a set of 20,000 galaxies classified by humans as part of Galaxy Cruise. The team then launched the AI on 700,000 galaxies in the Subaru Telescope data set.
The AI classified 400,000 of them as spiral galaxies and 30,000 as ring galaxies. Although ring galaxies represent less than 5% of all galaxies, this research yielded a large enough sample to conduct meaningful statistical analysis.
To continue reading, click here.