Saved in:
Bibliographic Details
Main Authors: Lu, Yao, Sun, Chunfeng, Xu, Dongwei, Lin, Yun, Xuan, Qi, Gui, Guan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.08300
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916944486072320
author Lu, Yao
Sun, Chunfeng
Xu, Dongwei
Lin, Yun
Xuan, Qi
Gui, Guan
author_facet Lu, Yao
Sun, Chunfeng
Xu, Dongwei
Lin, Yun
Xuan, Qi
Gui, Guan
contents Deep learning-based Automatic Modulation Recognition (AMR) model has made significant progress with the support of large-scale labeled data. However, when developing new models or performing hyperparameter tuning, the time and energy consumption associated with repeated training using massive amounts of data are often unbearable. To address the above challenges, we propose \emph{FoQuS}, which approximates the effect of full training by selecting a coreset from the original dataset, thereby significantly reducing training overhead. Specifically, \emph{FoQuS} records the prediction trajectory of each sample during full-dataset training and constructs three importance metrics based on training dynamics. Experiments show that \emph{FoQuS} can maintain high recognition accuracy and good cross-architecture generalization on multiple AMR datasets using only 1\%-30\% of the original data.
format Preprint
id arxiv_https___arxiv_org_abs_2509_08300
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle \emph{FoQuS}: A Forgetting-Quality Coreset Selection Framework for Automatic Modulation Recognition
Lu, Yao
Sun, Chunfeng
Xu, Dongwei
Lin, Yun
Xuan, Qi
Gui, Guan
Machine Learning
Artificial Intelligence
Deep learning-based Automatic Modulation Recognition (AMR) model has made significant progress with the support of large-scale labeled data. However, when developing new models or performing hyperparameter tuning, the time and energy consumption associated with repeated training using massive amounts of data are often unbearable. To address the above challenges, we propose \emph{FoQuS}, which approximates the effect of full training by selecting a coreset from the original dataset, thereby significantly reducing training overhead. Specifically, \emph{FoQuS} records the prediction trajectory of each sample during full-dataset training and constructs three importance metrics based on training dynamics. Experiments show that \emph{FoQuS} can maintain high recognition accuracy and good cross-architecture generalization on multiple AMR datasets using only 1\%-30\% of the original data.
title \emph{FoQuS}: A Forgetting-Quality Coreset Selection Framework for Automatic Modulation Recognition
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2509.08300