Some scientists believe that FRB comes from fast-rotating neutron stars, but their source has not yet been determined. This also explains the interesting thing about "Breakthrough Listening": Some people speculate that the explosion may be caused by intelligent aliens, or perhaps a spaceship is launched at an incredible speed in the universe. (Breakthrough Starshot, a sister project of Breakthrough Listen, is developing a laser-based optical navigation system, with the goal of launching micro detectors into the alien solar system in the next 30 years. [13 method of killing intelligent alien creatures]
FRB12102, which is about 3 billion light years away from the earth, is particularly attractive: it is only a known "repeater" source of FRB, otherwise it is often one-off.
In the new research, it is a breakthrough to listen to the team members of SETI (Searching for Alien Intelligence) Research Center at the University of California, Berkeley, applying machine learning technology to the data set in August of 20 17, which was originally analyzed by the Bank Telescope in Greenwest Virginia in a traditional way.
Researchers led by Gerry Zhang, a doctoral student at the University of California, Berkeley, trained an algorithm called convolutional neural network and found FRB in 400 megabytes of data. The representative of Breakthrough Listening said in a statement that this strategy is similar to that used by IT companies to optimize Internet search results.
Zhang and his colleagues discovered 72 bright spots, which made the total number of FRBs detected that day come from the same source (no matter what). By 1993,
"Not all the findings come from new observations," Peter Walden, executive director of Breakthrough, said in a statement. Breakthrough listening is part of a larger breakthrough initiative. It also has breakthrough scripts, breakthrough news and breakthrough observations. Worden added that in this case, it is intelligent and the original thinking is applied to the existing data set. It enhances our understanding of one of the most fascinating mysteries in astronomy.
Of course, this mystery still exists; We still don't know what FRBs is. However, Zhang said that the artificial intelligence method adopted in the new research may bring about various progress in the future.
"This work is just the beginning of using these powerful methods to find radio transients," he said in the same statement. We hope that our success will inspire others to make serious efforts to apply machine learning to radio astronomy.
This new paper has been accepted and published in the Journal of Astrophysics. You can read it for free on Breakthrough ListenFRB121102, and click Twitter@michaeldwall and Google+
Focus on Mike Wall. Follow us @Spacedotcom, Facebook or Google+. Originally published in space.