Convex Relaxations of ReLU Neural Networks Approximate Global Optima in Polynomial Time

Open in new window