1. Understanding the difficulty of training deep feedforward neural networks;Glorot,2010
2. Searching for activation functions;Ramachandran,2018
3. Parametric exponential linear unit for deep convolutional neural networks;Trottier,2017
4. TanhExp: a smooth activation function with high convergence speed for lightweight neural networks;Liu;IET Comput. Vis.,2021
5. Self-normalizing neural networks;Klambauer,2017