The Implicit Bias for Adaptive Optimization Algorithms on Homogeneous Neural Networks

Abstract

Despite their overwhelming capacity to overfit, deep neural networks trained by specific optimization algorithms tend to generalize well to unseen data. Recently, researchers explained it by investigating the implicit regularization effect of optimization algorithms. A remarkable progress is the work (Lyu&Li, 2019), which proves gradient descent (GD) maximizes the margin of homogeneous deep neural networks. Except GD, adaptive algorithms such as AdaGrad, RMSProp and Adam are popular owing to their rapid training process. However, theoretical guarantee for the generalization of adaptive optimization algorithms is still lacking. In this paper, we study the implicit regularization of adaptive optimization algorithms when they are optimizing the logistic loss on homogeneous deep neural networks. We prove that adaptive algorithms that adopt exponential moving average strategy in conditioner (such as Adam and RMSProp) can maximize the margin of the neural network, while AdaGrad that directly sums historical squared gradients in conditioner can not. It indicates superiority on generalization of exponential moving average strategy in the design of the conditioner. Technically, we provide a unified framework to analyze convergent direction of adaptive optimization algorithms by constructing novel adaptive gradient flow and surrogate margin. Our experiments can well support the theoretical findings on convergent direction of adaptive optimization algorithms.

arXiv commentICML 2021 Long Talk

References in corpus (15)

• Decoupled Weight Decay Regularization
• An overview of gradient descent optimization algorithms
• Recent Trends in Deep Learning Based Natural Language Processing
• The Loss Surfaces of Multilayer Networks
• Improving Generalization Performance by Switching from Adam to SGD
• On Exact Computation with an Infinitely Wide Neural Net
• A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks
• Adaptive Gradient Methods with Dynamic Bound of Learning Rate
• Characterizing Implicit Bias in Terms of Optimization Geometry
• Path-SGD: Path-Normalized Optimization in Deep Neural Networks
• Gradient descent aligns the layers of deep linear networks
• Gradient Descent Maximizes the Margin of Homogeneous Neural Networks
• Closing the Generalization Gap of Adaptive Gradient Methods in Training Deep Neural Networks
• Stochastic Gradient Descent on Separable Data: Exact Convergence with a Fixed Learning Rate
• When Will Gradient Methods Converge to Max-margin Classifier under ReLU Models?