Saved in:
Bibliographic Details
Main Authors: Jin, Long, Nong, Han, Chen, Liangming, Su, Zhenming
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.12473
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916527123464192
author Jin, Long
Nong, Han
Chen, Liangming
Su, Zhenming
author_facet Jin, Long
Nong, Han
Chen, Liangming
Su, Zhenming
contents The insufficient generalization of adaptive moment estimation (Adam) has hindered its broader application. Recent studies have shown that flat minima in loss landscapes are highly associated with improved generalization. Inspired by the filtering effect of integration operations on high-frequency signals, we propose multiple integral Adam (MIAdam), a novel optimizer that integrates a multiple integral term into Adam. This multiple integral term effectively filters out sharp minima encountered during optimization, guiding the optimizer towards flatter regions and thereby enhancing generalization capability. We provide a theoretical explanation for the improvement in generalization through the diffusion theory framework and analyze the impact of the multiple integral term on the optimizer's convergence. Experimental results demonstrate that MIAdam not only enhances generalization and robustness against label noise but also maintains the rapid convergence characteristic of Adam, outperforming Adam and its variants in state-of-the-art benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2412_12473
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Method for Enhancing Generalization of Adam by Multiple Integrations
Jin, Long
Nong, Han
Chen, Liangming
Su, Zhenming
Machine Learning
The insufficient generalization of adaptive moment estimation (Adam) has hindered its broader application. Recent studies have shown that flat minima in loss landscapes are highly associated with improved generalization. Inspired by the filtering effect of integration operations on high-frequency signals, we propose multiple integral Adam (MIAdam), a novel optimizer that integrates a multiple integral term into Adam. This multiple integral term effectively filters out sharp minima encountered during optimization, guiding the optimizer towards flatter regions and thereby enhancing generalization capability. We provide a theoretical explanation for the improvement in generalization through the diffusion theory framework and analyze the impact of the multiple integral term on the optimizer's convergence. Experimental results demonstrate that MIAdam not only enhances generalization and robustness against label noise but also maintains the rapid convergence characteristic of Adam, outperforming Adam and its variants in state-of-the-art benchmarks.
title A Method for Enhancing Generalization of Adam by Multiple Integrations
topic Machine Learning
url https://arxiv.org/abs/2412.12473