A group of scientists from the University of Georgia have developed an innovative approach for detecting and classifying new worlds far from Earth. They discovered an exoplanet by utilising machine learning, a branch of artificial intelligence (AI).
A recent study demonstrated that AI can detect the presence of an exoplanet by studying protoplanetary discs, the gas surrounding freshly formed stars. The recently released data provide a first step towards using machine learning to identify previously overlooked exoplanets.
Jason Terry, who is the Lead author of the study, said: “We confirmed the planet using traditional techniques, but our models directed us to run those simulations and showed us exactly where the planet might be.
“When we applied our models to a set of older observations, they identified a disk that wasn’t known to have a planet despite having already been analyzed. Like previous discoveries, we ran simulations of the disk and found that a planet could re-create the observation.”
Terry added that the models showed the presence of a planet by taking repeated images of a specific section of the disc that turned out to contain the precise sign of a planet – an exceptional change in the velocity of the gas surrounding the planet.
Meanwhile, the assistant professor of computational astrophysics and principal investigator of the Exoplanet and Planet Formation Research Group at UGA, Cassandra Hall, stated, “This is an incredibly exciting proof of concept.
We knew from our previous work that we could use machine learning to find known forming exoplanets. Now, we know for sure that we can use it to make brand discoveries.
“This demonstrates that our models – and machine learning in general – have the ability to quickly and accurately identify important information that people can miss. This has the potential to dramatically speed up analysis and subsequent theoretical insights,” the author noted.
“It only took about an hour to analyze that entire catalogue and find strong evidence for a new planet in a specific spot, so we think there will be an important place for these types of techniques as our datasets get even larger.”
?