Scientists have developed a machine learning method, which can researchers filter out interference and more effectively detect unusual radio signals from space, contributing to the search for an extraterrestrial mind.
The programs for the search for extraterrestrial mind (SETI) have used radio telescopes for decades to detect space signals. However, this search was complicated due to the development of technologies that can generate false positive identities that require a large amount of time to filter from large data sets.
SETI scientists simulated many signals, introduced them into real observations and taught the ML system to classify these simulations. Part of the autoencoder studies both on real observations and simulations when reconstructing the initial input data, which helps specialists to distinguish the characteristic features of the input image. Together, this helps to create an effective algorithm for detecting anomalies.
in research under the guidance of Peter Ma, a 3rd year student of the Toronto University of Physics and Mathematics, Observations of 820 stars in the form of 115 million data fragments were used. The ML models developed by the TensorFlow machine learning library and Python Keras libraries revealed about 3 million. Then the set was reduced to 20,515 signals, which is more than 100 times less than in the previous analysis of the same data set.
Scientists revealed 8 previously undetected signals, although in subsequent observations these signals were not displayed. The authors suggest that their method can be applied to other large data sets to accelerate SETI and similar polls based on data.
Other developed ML applications in Seti include:
- General classifier of signals for observations obtained on an array of allen telescopes and on a spherical radio telescope with a 500-meter aperture;
- Radio frequency interference identifiers based on adulthood neural networks;
- the detection algorithms .
“Seti seeks to find evidence of a reasonable life in space with the help of” technology “created by scientists. The detection of unambiguous technology will demonstrate the existence of an extraterrestrial mind (ETI), therefore this is of great interest for both scientists and the general public,” – The researchers said.