Far-reaching radio bursts, perhaps from an advanced civilization, identified with the help of a new research method | New

(RNN) – The new technology has allowed researchers at the Berkeley SETI Research Center to identify distant radio emissions that could be "technology signatures developed by an advanced civilization".

UC Berkeley's "Breakthrough Listen" program used machine learning to identify 72 new recordings known as fast radio bursts from a strange repetitive burst called FRB 121102.

Gusts usually last for milliseconds. Depending on the program, the burst can emit as much energy in 10 microseconds as the sun in the whole year.

As the program explains, "most FRBs were observed during a single explosion," while this signal "is the only one so far known to generate repeated gusts."

According to the group, this included 21 observations from last year's breakup. The new set of 72 was found by examining old datasets.

Machine learning is a process by which, as a rule, computers are programmed to take huge sets of data and "learn" to classify them and organize them into practical information.

"All the discoveries do not come from new observations," said Pete Worden, executive director of the program, in the statement. "In this case, it was an intelligent and original reflection applied to an existing data set. He advanced our knowledge of one of the most enticing mysteries of astronomy. "

The source of recently discovered radio bursts is a galaxy located about 3 billion light years from Earth.

The FRB 121102 was observed for the first time in 2012 and in 2015 it was found that it was repeated.

As reported in the New York Times in January, "Among the most common explanations were lasers propelling interstellar spaceships."

Less extraordinary possibilities could explain the splinters. The publication also noted a theory that they could have come from "highly magnetized neutron stars" (essentially a collapsed form of a once massive star). Other more conservative analyzes have suggested this theory.

The researchers used an algorithm known as the "convolutional neural network" to process a massive set of data, 400 terabytes, and identify the bursts missed during last year's search.

One terabyte of data equates to approximately 310,000 images.

Andrew Siemion, director of the Berkeley SETI Research Center, said in an article by UC Berkeley that successful application of machine learning could lead to further advances.

"This work is exciting, not only because it helps us better understand the dynamic behavior of fast radio bursts," he said. "But also because of the promise made for the use of machine learning to detect signals missed by conventional algorithms."

Copyright 2018 Raycom News Network. All rights reserved.