Convex Formulations for Training Two-Layer ReLU Neural Networks.

Karthik Prakhya, Tolga Birdal, Alp Yurtsever

Browse the full ICLR paper archive.