In a recent study published in the journal Nature Astronomy, a new machine-learning algorithm has been used to detect eight extraterrestrial signals that appear to be of technological origin. The research team, led by Peter Ma, an undergraduate at the University of Toronto, used a technique known as “semi-unsupervised learning” which combines supervised and unsupervised machine learning.
The researchers analyzed 150 terabytes of data from the Green Bank Telescope in West Virginia, covering observations of 820 stars near Earth. They discovered eight previously overlooked signals from five stars located between 30 light-years and 90 light-years from Earth. The signals were found to have two common features with signals that could be created by intelligent aliens.
Using Machine Learning to Search for Extraterrestrial Intelligence
The researchers first trained the algorithm to differentiate between human-caused signals from radio waves on Earth and radio signals from elsewhere. The team tested different algorithms to minimize false positives and used the Green Bank Telescope in West Virginia to analyze the data.
The study co-author, Cherry Ng, an astronomer at the University of Toronto, expressed optimism in using artificial intelligence in the search for extraterrestrial signals. She stated, “I’m optimistic that we’ll be able to better quantify the likelihood of the presence of extraterrestrial signals from other civilizations.”
Next Steps in the Search for Extraterrestrial Life
While the signals detected in the study do not provide proof of intelligent aliens, they do represent a promising step in the search for extraterrestrial life. The researchers plan to apply their algorithm to data from powerful radio telescopes, such as MeerKAT in South Africa and the planned Next Generation Very Large Array, which will be distributed across North America.
The research team hopes that their new technique, combined with the next generation of telescopes, will allow them to search millions of stars for extraterrestrial signals. Peter Ma said, “With our new technique, combined with the next generation of telescopes, we hope that machine learning can take us from searching hundreds of stars to searching millions.”