Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

Machine learning program tracks down mysterious cosmic radio bursts

A new machine learning program discovered radio bursts from a mysterious source 3 billion light years from Earth, according to new research from the Berkeley SETI Research Centre at the University of California.

The artificial intelligence (AI) algorithm is part of Breakthrough Listen, a science-based program developed in 2015 to search for intelligent extra-terrestrial communications in the Universe, and researchers believe it might have intercepted signals from an advanced civilisation on a far-away planet.

Fast radio bursts are bright pulses of radio emission mere milliseconds in duration, thought to originate from distant galaxies. Although the source of these emissions is still unclear, most fast radio bursts are one-offs.

The source examined by astronomers in this instance though is unique in emitting repeated bursts, with over 300 bursts detected since it was discovered in 2012. This unusual behaviour has drawn the attention of many researchers hoping to pin down the cause and origin of the signal.

The algorithm that analysed the signal was trained by a neural network to recognise bursts found by classical search methods, and then re-analysed the dataset to find bursts that the classical approach missed.

“This work is exciting not just because it helps us understand the dynamic behaviour of fast radio bursts in more detail, but also because of the promise it shows for using machine learning to detect signals missed by classical algorithms,” said Andrew Siemion, director of the Berkeley SETI Research Center and principal investigator for Breakthrough Listen.

“Whether or not FRBs themselves eventually turn out to be signatures of extra-terrestrial technology, Breakthrough Listen is helping to push the frontiers of a new and rapidly growing area of our understanding of the Universe around us,” he added.

The results of the research have been published in The Astrophysical Journaland, and they’re available for download from the Breakthrough Listen website.


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter