This afternoon we have a call with the gravitational wave specialists in the Einstein Telescope A.I. team on a new idea that is running on the side of Boosting Alpha. This creative idea was inspired when one of the top astrologists mentioned that he was listening to the signals coming out of the existing GW telescopes 10 years ago with his earphones on. The human ears are still superior to machines.

This inspired us on the hypothesis to use A.I. techniques from the music industry (the same techniques that Spotify uses to learn which type of music you like) and to try to see if it works on gravitational wave signals. So what we are trying to pilot now is to:
– take the raw signals coming from the GW telescopes
– calculate 50 to 100 different input features inspired from libraries used in the music industry (in our case Librosa and Essentia)
– build an A.I. model that uses these input features and tries to learn the correlation with the events (black hole mergers, neutron stars mergers) that were already found with GW detectors in the last 5 years
– later we also want to test whether the A.I. model can accurately distinguish between real events (which happen on average 1 time per week) and so-called glitches (which happen on average 1 time per second and are caused by factors on earth).

Anxious to experience what the GW experts think of this crazy idea.
#boostingalpha #einsteintelescope #gravitationalwaves