An international research team has developed a sophisticated deep learning algorithm that combines Graphics Processing Unit (GPU) phase folding with neural networks. By using this innovative approach, the researchers identified five ultra-short period planets, each with a diameter smaller than Earth's and an orbital period under 24 hours. This algorithm utilized data from the Kepler space telescope's 2017 release.
Among these newly discovered celestial bodies, four are the smallest exoplanets ever found, located at the closest proximity to their respective host stars, with sizes comparable to Mars.
This achievement marks the first time AI has been employed by astronomers for both searching for candidate signals and pinpointing genuine ones. The study's findings have been published in the Monthly Notices of the Royal Astronomical Society (MNRAS).
The research is led by Professor Ge Jian of the Shanghai Astronomical Observatory, Chinese Academy of Sciences.
Reference(s):
cgtn.com
