重庆理工大学学报2025，Vol.39Issue(9)：98-106,9.DOI:10.3969/j.issn.1674-8425(z).2025.05.012
多目标优化灰狼算法改进长短期记忆网络的睡眠分期研究
Research on sleep staging improvement of long short-term memory network using multi-objective grey wolf algorithm
摘要
Abstract
To address the low accuracy and the possible misjudgment in the N1-stage of sleep staging,this study proposes a multi-objective Grey Wolf Optimizer-enhanced LSTM model(DE-GWO-LSTM).The framework integrates differential evolution(DE)into Grey Wolf Algorithm(GWA)to optimize LSTM hyper-parameters,dynamically adjusting hidden layer nodes and preventing local optima through position-update refinement.Representative time-frequency and nonlinear features extracted from the ISRUC-Sleep dataset are processed by the model for stage classification.Experimental results demonstrate the proposed model achieves an overall accuracy of 88.6%,N1-stage precision of over 70%,outperforming existing models.关键词
睡眠分期/灰狼算法/长短期记忆网络/差分算法Key words
sleep staging/GWA/LSTM/DE分类
计算机与自动化引用本文复制引用
高鹏强,丁顺良,宛磊,李奎,吴广良,高建设..多目标优化灰狼算法改进长短期记忆网络的睡眠分期研究[J].重庆理工大学学报,2025,39(9):98-106,9.基金项目
国家自然科学基金项目(51906225) （51906225）
中国国家留学基金项目(202308410392) （202308410392）
