Understanding the role of (stochastic) gradient descent in training and generalisation of deep neural networks with ReLU activation has been the object study in the recent past. In this paper, we make use of deep gated networks (DGNs) as a framework to obtain insights about DNNs with ReLU activation. In DGNs, a single neuronal unit has two components namely the pre-activation input (equal to the inner product the weights of the layer and previous layer outputs), and a gating value which belongs to $[0,1]$ and the output of the neuronal unit is equal to the multiplication of pre-activation input and the gating value. The standard DNN with ReLU gate, is a special case of the DGNs, wherein the gating value is $1/0$ based on whether or not the pre-activation input is positive or negative. We theoretically analyse and experiment with several variants of DGNs, each variant suited to understand a particular aspect of either training/generalisation in DNNs with ReLU activations. Our theory throws light on two questions namely i) why increasing depth till a point helps in training and ii) why increasing depth beyond a point hurts training? We also present experimental evidence to show that gate adaptation is key for generalisation.

通过深门控网络作为框架，理论分析和实验验证了几种DGNs变体，阐述了深度神经网络的训练和泛化方面的问题，着重讨论了为什么增加深度能帮助训练到一定程度，而增加深度会伤害训练，以及证明了门控自适应在泛化中的关键作用。

深层门控网络：深度学习训练和泛化的框架理解