Towards Understanding Regularization in Batch Normalization
This work provides theoretical insights into a widely used technique in deep learning, but it is incremental as it builds on existing understanding of regularization effects.
The authors tackled the problem of understanding why Batch Normalization improves neural network training by analyzing it as an implicit regularizer, decomposing it into population normalization and gamma decay, and showing that training converges with large learning rates. Experiments confirmed these regularization traits in convolutional neural networks.
Batch Normalization (BN) improves both convergence and generalization in training neural networks. This work understands these phenomena theoretically. We analyze BN by using a basic block of neural networks, consisting of a kernel layer, a BN layer, and a nonlinear activation function. This basic network helps us understand the impacts of BN in three aspects. First, by viewing BN as an implicit regularizer, BN can be decomposed into population normalization (PN) and gamma decay as an explicit regularization. Second, learning dynamics of BN and the regularization show that training converged with large maximum and effective learning rate. Third, generalization of BN is explored by using statistical mechanics. Experiments demonstrate that BN in convolutional neural networks share the same traits of regularization as the above analyses.