Training deep neural networks with gradual deconvexification.

James Ting-Ho Lo, Yichuan Gui, Yun Peng

Browse the full IJCNN paper archive.