Scientists use artificial intelligence to locate the planet outside our solar system

A group of scientists from the University of Georgia has created a new way to discover and classify new worlds far from Earth. They encountered an exoplanet using machine learning, a suite of artificial intelligence (AI).

The latest scan showed that AI can accurately identify the presence of an exoplanet by examining protoplanetary disks, the gas around newly created stars. The newly released results mark the first step toward using machine learning to identify exoplanets that have previously been missed.

“We confirmed the planet using conventional techniques, but our models directed us to run those simulations and showed us exactly where the planet might be,” said lead author of the study Jason Terry in a letter.

“When we applied our models to a group of ancient observations, they identified a disk that was not known to contain a planet even though it had already been analyzed. Like previous discoveries, we ran simulations of the disk and found that the planet could be a planet-creating observation.”

According to Terry, the models detected a planet, marked by multiple images that strongly emphasize a specific portion of the disk that turned out to have a specific indicator of a planet — an unusual variation in the velocity of gas around the planet.

“This is an incredibly exciting proof of concept,” said Cassandra Hall, assistant professor of computational astrophysics and principal investigator in the Exoplanet and Planet Formation Research Group at UGA. “We knew from our previous work that we can use machine learning to find known-forming exoplanets.” Now, we know for sure that we can use it to discover brands.

“This shows that our models – and machine learning in general – have the ability to quickly and accurately identify critical information that people might miss. This has the potential to dramatically speed up analysis and subsequent theoretical insights,” said Terry.

“It took about an hour to analyze this entire catalog and find strong evidence for a new planet in a particular location, so we think there will be an important place for these types of technologies as our data sets get larger.”