Non-convergence to the optimal risk for Adam and stochastic gradient descent optimization in the training of deep neural networks

Re-assign community