Fig. 3From: Convergence of batch gradient learning with smoothing regularization and adaptive momentum for neural networksThe curve of norm of gradientBack to article page