Gradient-Variation Online Learning under Generalized Smoothness

Authors: Yan-Feng Xie, Peng Zhao, Zhi-Hua Zhou

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this paper, we systematically study gradient-variation online learning under generalized smoothness. We extend the classic optimistic mirror descent algorithm to derive gradient-variation regret by analyzing stability over the optimization trajectory and exploiting smoothness locally. Then, we explore universal online learning, designing a single algorithm with the optimal gradient-variation regrets for convex and strongly convex functions simultaneously, without requiring prior knowledge of curvature. This algorithm adopts a two-layer structure with a meta-algorithm running over a group of base-learners. To ensure favorable guarantees, we design a new Lipschitz-adaptive meta-algorithm, capable of handling potentially unbounded gradients while ensuring a second-order bound to effectively ensemble the base-learners. Finally, we provide the applications for fast-rate convergence in games and stochastic extended adversarial optimization.
Researcher Affiliation Academia Yan-Feng Xie, Peng Zhao, Zhi-Hua Zhou National Key Laboratory for Novel Software Technology, Nanjing University, China School of Artificial Intelligence, Nanjing University, China {xieyf, zhaop, zhouzh}@lamda.nju.edu.cn
Pseudocode Yes Algorithm 1 Lipschitz Adaptive Optimistic Adapt-ML-Prod; Algorithm 2 Universal Gradient-Variation Online Learning under Generalized Smoothness
Open Source Code No This paper does not include experiments, and no data or code will be provided.
Open Datasets No This paper does not include experiments.
Dataset Splits No This paper does not include experiments.
Hardware Specification No This paper does not include experiments.
Software Dependencies No This paper does not include experiments.
Experiment Setup No This paper does not include experiments.