On the Double Descent of Random Features Models Trained with SGD

Open in new window