Researchers employed machine learning technologies, which are similar to artificial intelligence, to validate the existence of previously unknown exoplanets. According to the researchers from the University of Georgia in the United States, their work demonstrated that machine learning can correctly predict the presence of exoplanets by searching in protoplanetary disks, the gas surrounding freshly formed stars.
They stated that their findings, which were published in the Astrophysical Journal, were a first step toward utilizing machine learning to identify previously unknown exoplanets. “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,” said Jason Terry, a doctoral student in the Department of Physics and Astronomy at the University of Georgia and the study’s lead author.
“When we applied our models to a set of older observations, they identified a disc that wasn’t known to have a planet despite having already been analyzed. Like previous discoveries, we ran simulations of the disc and found that a planet could recreate the observation” said Terry.
According to Terry, the models showed the presence of a planet, as evidenced by many photographs that clearly highlighted a specific section of the disk that turned out to contain the classic sign of a planet–an unexpected deviation in the velocity of the gas surrounding the planet.
“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’ said Cassandra Hall, assistant professor of computational astrophysics and principal investigator of the Exoplanets and Planet Formation Research Group at UGA. “Now, we know for sure that we can use it to make brand new discoveries,’ said Hall.
Machine Learning Algorithms uncover previously undetected signals in Data Analysis
The algorithms were able to detect a signal in previously analyzed data; they discovered something that had previously gone undiscovered. According to the researchers, their study demonstrates how using Al might improve scientists’ jobs by increasing accuracy and economically economizing time. (editorialrm.com)
“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,” Terry said.
