AI finds signals eliminating galaxies

SETI (Search for Extraterrestrial Intelligence) has recently deployed a custom neural network over its huge dataset, which leads to the discovery of a galaxy emitting 72 fast radio bursts at a distance of about 3 billion miles.

As this new comes from SETI, you might be thinking it as a sign of extraterrestrial life trying to contact us, right?

But as it turns out, it isn't. In fact, it is a cosmic phenomenon shown by certain celestial bodies which are named FRB(for fast radio burst). This time it was FRB 121102, which is the only known object to emit it regularly. These signals were observed by Green Bank Telescope in West Virginia, which was pointed toward this source of fast and bright (hence the name) bursts for five hours in August of 2017 and size of data collected was 400 terabytes.

So find this signal was quite tough and originally only 21 FRB's were identified. But recently as a part of Breakthrough project with SETI, Gerry Zhang a student of UC Berkeley has developed a neural network for this tough job. And results are outstanding and 72 other FRB's were identical. But, no pattern long than 10 milliseconds can be found, chances of it being from extraterrestrial is next to none.

Whatever the case, SETI and Breakthrough will continue to observe FRB 121102 for a while for new discoveries.
Paper on this discovery, co-authored by Cal postdoc Vishal Gajjar, is yet to be published in the Astrophysical Journal.

“Gerry’s work is exciting not just because it helps us understand the dynamic behavior of FRBs in more detail, but also because of the promise it shows for using machine learning to detect signals missed by classical algorithms,” explained Berkeley’s Andrew Siemion, who leads the SETI research.