On the Hardness of Learning One Hidden Layer Neural Networks

Open in new window