← Search

Sheng Long

2 accepted papers

2024

On the Convergence of an Adaptive Momentum Method for Adversarial Attacks

AAAI 2024technical

Adversarial examples are commonly created by solving a constrained optimization problem, typically using sign-based methods like Fast Gradient Sign Method (FGSM). These attacks can benefit from momentum with a constant parameter, such as Momentum Iterative FGSM (MI-FGSM), to enhance black-box transf…

Cited by 6SourcePDFScholar
2021

The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods

ICLR 2021poster

The adaptive stochastic gradient descent (SGD) with momentum has been widely adopted in deep learning as well as convex optimization. In practice, the last iterate is commonly used as the final solution. However, the available regret analysis and the setting of constant momentum parameters only guar…

Cited by 17SourcePDFScholar