Artificial intelligence (AI) has also been shown to be useful in identifying extraterrestrial signals.
US Vice reported on the 31st (local time) that researcher Peter Ma and others at the SETI Research Center at Berkeley University in the US analyzed 480 hours of data collected from 820 stars with radio telescopes using machine learning algorithms. Results reported finding eight previously unidentified signals.
The signal found this time was a Doppler drift signal with a narrow bandwidth and had a specific frequency. However, it was not yet known how these signals change with time and distance.
Research has yet to be done to determine what technology is connected to the signal. Best case scenario, you’ll find embedded technological information or traces of extraterrestrial civilizations in these signals.
Since its establishment in 1984, the SETI Institute has been working on finding extraterrestrial radio signals. Scientists looking for traces of extraterrestrial life are paying attention to radio signals because they can be evidence of the existence of communication technology and the intelligence that creates them.
However, in the past few decades, little progress has been made. Radio data floating in space easily mixes with signals from Earth, and finding signals from outer space is like finding a needle in a haystack.
While the SETI Institute has introduced elements of machine learning to these search tasks, it has maintained a human-supervised structure. On the other hand, researchers such as Peter Ma and others found eight signals as a result of choosing a method of completely entrusting the work to the algorithm.
“The algorithm finds only what we tell it to look for, but the problem is that the nature of the extraterrestrial signal is unknown,” said Peter Ma. “The approach we proposed was to learn from it.”
Jeong Byeong-il, member jbi@aitimes.com