Exoplanets, or planets that orbit stars other than the Sun, have been a fascination for the scientific community for decades. The search for these planets has been a key goal in space exploration, and their discovery has been made possible by technological advances in astronomical observation.
Now, a team of researchers from the University of Georgia (UGA) has used machine learning to discover a previously unknown exoplanet.
New planetary discovery thanks to machine learning technology
The study behind this revelation was conducted by UGA Franklin College of Arts and Sciences physics and sciences doctoral student Jason Terry and assistant professor of computational astrophysics Cassandra Hall. Using machine learning tools, the team identified a protoplanetary disk that had not previously been analyzed for the presence of an exoplanet.
The discovery was made possible because the models used by the team were able to detect a signal in data that had been previously analyzed, but that people had missed. These models can dramatically speed up subsequent analysis and theoretical thinking, allowing scientists to more efficiently economize their time when engaged in deep outer space research.
The use of machine learning in exoplanet detection is an important step in space exploration. Scientists have been using traditional techniques to detect exoplanets, but machine learning’s ability to find previously overlooked exoplanets is a breakthrough. This opens the door to the possibility of finding more planets in the future.
“We confirmed the planet using traditional techniques, but our models directed us to run those simulations and showed us exactly where the planet could be,” Jason Terry said. “When we applied our models to a set of older observations, they identified a disk that was not 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 recreate the observation.”he added.
Machine learning technology is used for the analysis of large data sets, addressing this challenge that has historically been difficult for scientists in traditional settings. In addition, the ability of machine learning to detect signals that may have been missed by human scientists provides valuable additional help.
In the case of this work, the researchers say that it only took them about an hour to go through that entire catalog and find strong evidence for a new planet in a specific location, so they believe that there will be an important place for this type of technique to come. as our data sets get even bigger.
