Fetching the paper…

Overall error analysis for the training of deep neural networks via stochastic gradient descent with random initialisation · Around