PLANET-NOONS: pipeline for analyzing new exoplanet transits using neural networks to operate and observe universal surveys

Joshua Zou, California, USA 16-18

Project PLANET-NOONS is a neural network able to self-update quickly without losing accuracy, cutting training times by up to 90% versus the previous most-accurate neural network applied in the field of astronomy. This self-updating prototype can also be applied to multiple different observational projects at the same time, and can pull and compile knowledge from numerous professional and civilian research projects, greatly improving the model’s ability to deal with diverse issues such as increased distortion or eliminating background events. Finally, as a proof of process, I identified and confirmed 26 new exoplanet candidates with the neural network.

Show more