Seismology with Adversarial Recurrent Stars
Can we use Generative Adversarial Networks to model red giant light curves, and then use the GANs to model their noise and look for planets?
We've got a thousand of the best and brightest red giants, and we're going to train our GAN on one-day segments of long-cadence data.
This is a project developed for TESS Ninja 2 held at the University of Chicago, March 25-29 2019.