Statistical Physics of Deep Neural Networks: Initialization toward Optimal Channels

Open in new window