Optimal Rates for Generalization of Gradient Descent for Deep ReLU Classification

Open in new window