Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: You, Bihao, Cui, Jiping
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2512.12273
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915673180995584
author You, Bihao
Cui, Jiping
author_facet You, Bihao
Cui, Jiping
contents Prediction of epilepsy based on electroencephalogram (EEG) signals is a rapidly evolving field. Previous studies have traditionally applied 1D processing to the entire EEG signal. However, we have adopted the Gram Matrix method to transform the signals into a 3D representation, enabling modeling of signal relationships across dimensions while preserving the temporal dependencies of the one-dimensional signals. Additionally, we observed an imbalance between local and global signals within the EEG data. Therefore, we introduced multi-level feature extraction, utilizing coattention for capturing global signal characteristics and an inception structure for processing local signals, achieving multi-granular feature extraction. Our experiments on the BONN dataset demonstrate that for the most challenging five-class classification task, GRC-Net achieved an accuracy of 93.66%, outperforming existing methods.
format Preprint
id arxiv_https___arxiv_org_abs_2512_12273
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GRC-Net: Gram Residual Co-attention Net for epilepsy prediction
You, Bihao
Cui, Jiping
Machine Learning
Artificial Intelligence
Prediction of epilepsy based on electroencephalogram (EEG) signals is a rapidly evolving field. Previous studies have traditionally applied 1D processing to the entire EEG signal. However, we have adopted the Gram Matrix method to transform the signals into a 3D representation, enabling modeling of signal relationships across dimensions while preserving the temporal dependencies of the one-dimensional signals. Additionally, we observed an imbalance between local and global signals within the EEG data. Therefore, we introduced multi-level feature extraction, utilizing coattention for capturing global signal characteristics and an inception structure for processing local signals, achieving multi-granular feature extraction. Our experiments on the BONN dataset demonstrate that for the most challenging five-class classification task, GRC-Net achieved an accuracy of 93.66%, outperforming existing methods.
title GRC-Net: Gram Residual Co-attention Net for epilepsy prediction
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2512.12273