Regularizing Neural Networks via Adversarial Model Perturbation
Effective regularization techniques are highly desired in deep learning for alleviating overfitting and improving generalization. This work proposes a new regularization scheme, based on the understanding that the flat local minima of the empirical risk cause the model to generalize better. This scheme is referred to as adversarial model perturbation (AMP), where instead of directly minimizing the empirical risk, an alternative "AMP loss" is minimized via SGD. Specifically, the AMP loss is obtained from the empirical risk by applying the "worst" norm-bounded perturbation on each point in the parameter space. Comparing with most existing regularization schemes, AMP has strong theoretical justifications, in that minimizing the AMP loss can be shown theoretically to favour flat local minima of the empirical risk. Extensive experiments on various modern deep architectures establish AMP as a new state of the art among regularization schemes. Our code is available at https://github.com/hiyouga/AMP-Regularizer.
Code (1)
Tasks
Image ClassificationmodelMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Improving Resistance to Adversarial Deformations by Regularizing Gradients
Improving the resistance of deep neural networks against adversarial attacks is important for deploying models to realistic applications. However, most defense methods are designed to defend against intensity perturbatio…
δ-SAM: Sharpness-Aware Minimization with Dynamic Reweighting
Deep neural networks are often overparameterized and may not easily achieve model generalization. Adversarial training has shown effectiveness in improving generalization by regularizing the change of loss on top of adve…
Sharpness-Aware Minimization with Dynamic Reweighting
Deep neural networks are often overparameterized and may not easily achieve model generalization. Adversarial training has shown effectiveness in improving generalization by regularizing the change of loss on top of adve…
Natural Language UnderstandingImproving Back-Propagation by Adding an Adversarial Gradient
The back-propagation algorithm is widely used for learning in artificial neural networks. A challenge in machine learning is to create models that generalize to new data samples not seen in the training data. Recently, a…
BIG-bench Machine LearningGeneral ClassificationAdversarial Training Methods for Semi-Supervised Text Classification
Adversarial training provides a means of regularizing supervised learning algorithms while virtual adversarial training is able to extend supervised learning algorithms to the semi-supervised setting. However, both metho…
ClassificationGeneral ClassificationSemi-Supervised Text ClassificationSentiment Analysis+2