Machine learning could be a key tool in the search for extraterrestrial life. Recent researches, all led by a University of Toronto undergraduate student and supported by the SETI Institute, have discovered and investigated eight previously unidentified signals. How does machine learning help in the search for extraterrestrial life?
In this Befree blog post we discuss the discoveries made by the student, Peter Ma. How does machine learning help in the search for extraterrestrial life?
What is machine learning?
Machine learning is a branch of artificial intelligence that deals with developing algorithms and techniques that allow computers to learn from data and improve their performance without being explicitly programmed. In other words, the goal is to enable computers to perform complex tasks without detailed, situation-specific programming. How does machine learning help in the search for extraterrestrial life?
Its origins lie in the 1950s, when the first concepts and theories related to machine learning were developed. However, it was not until the 1990s that it began to take shape and become a discipline in its own right, thanks to the growing capacity of data processing and storage.
Today, the applications of machine learning are varied. They range from automating business processes to improving the user experience in mobile applications. Examples include classifying emails as spam or non-spam, personalising online recommendations, detecting credit card fraud, improving energy efficiency in factories and improving accuracy in medical diagnostics. In addition, we could now target machine learning to find extraterrestrial life.
AI in search of extraterrestrials
The study re-examined data taken with the Green Bank telescope in West Virginia as part of a Breakthrough Listen campaign. The goal was to apply new deep learning techniques to a classic search algorithm to improve the results. After running the new algorithm and manually reviewing the data, the newly detected signals had several interesting features, such as reduced spectral width, non-zero drift rates, and occurrence only in ON-source observations.
This study demonstrates the importance of using machine learning techniques in the search for extraterrestrial life. Classical algorithms can be outdated and inefficient with modern datasets, while deep learning techniques allow processing large amounts of information more quickly and accurately. Moreover, the application of these techniques to a dataset previously considered devoid of interesting signals has resulted in a surprising discovery.
This finding also underscores machine learning’s ability to revolutionise the way we investigate the universe. With the ability to process vast amounts of data and detect patterns and trends that may go unnoticed by human eyes, machine learning opens the door to new forms of exploration and discovery in a wide range of fields, including astrophysics. It is likely that in the future we will see more applications of machine learning in the search for extraterrestrial life, increasing our chances of finding interesting signals and discovering new worlds.
Nothing definitive yet
Despite the magnitude of the discoveries, it is important to keep in mind that these signals are not necessarily evidence of extraterrestrial life. They may be caused by natural phenomena or even errors in equipment or data. Still, these results suggest that there may be more to discover in the search for extraterrestrial life and that the combination of classical and machine learning techniques may be the key to making new discoveries.
Moreover, machine learning alone cannot guarantee the discovery of extraterrestrial life. The search for extraterrestrial intelligence is an interdisciplinary effort involving experts in astronomy, planetary science, biology, philosophy and other fields. Machine learning is only one valuable tool in the search for technosignatures, and its success depends on integration with other techniques and expert interpretation.
