《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (5): 1445-1451.DOI: 10.11772/j.issn.1001-9081.2023050747

• 2023年中国计算机学会人工智能会议(CCFAI 2023) • 上一篇    

小样本场景下的元迁移学习睡眠分期模型

时旺军1,2, 王晶1,2,3(), 宁晓军1,2, 林友芳1,2,3   

  1. 1.北京交通大学 计算机科学与技术学院, 北京 100044
    2.交通数据分析与挖掘北京市重点实验室(北京交通大学), 北京 100044
    3.民航旅客服务智能化应用技术重点实验室(中国民用航空局), 北京 101318
  • 收稿日期:2023-06-09 修回日期:2023-07-05 接受日期:2023-07-09 发布日期:2023-08-01 出版日期:2024-05-10
  • 通讯作者: 王晶
  • 作者简介:时旺军(1999—),男,天津人,硕士研究生,主要研究方向:数据挖掘、机器学习
    宁晓军(1999—),男,山东泰安人,博士研究生,CCF会员,主要研究方向:脑机接口、时间序列分析与挖掘
    林友芳(1971—),男,福建武平人,教授,博士,CCF高级会员,主要研究方向:智能系统、复杂网络、交通数据挖掘。
    第一联系人:王晶(1987—),女,安徽合肥人,副教授,博士,CCF会员,主要研究方向:时间序列分析与挖掘
  • 基金资助:
    中央高校基本科研业务费专项资金资助项目(2023JBMC056)

Sleep stage classification model by meta transfer learning in few-shot scenarios

Wangjun SHI1,2, Jing WANG1,2,3(), Xiaojun NING1,2, Youfang LIN1,2,3   

  1. 1.School of Computer Science and Technology,Beijing Jiaotong University,Beijing 100044,China
    2.Beijing Key Lab of Traffic Data Analysis and Mining (Beijing Jiaotong University),Beijing 100044,China
    3.Key Laboratory of Intelligent Application Technology for Civil Aviation Passenger Services (Civil Aviation Administration of China),Beijing 101318,China
  • Received:2023-06-09 Revised:2023-07-05 Accepted:2023-07-09 Online:2023-08-01 Published:2024-05-10
  • Contact: Jing WANG
  • About author:SHI Wangjun, born in 1999, M. S. candidate. His research interests include data mining, machine learning.
    NING Xiaojun, born in 1999, Ph. D. candidate. His research interests include brain computer interface, time series analysis and mining.
    LIN Youfang, born in 1971, Ph. D., professor. His research interests include intelligent systems, complex networks, traffic data mining.
  • Supported by:
    Fundamental Research Funds for Central Universities(2023JBMC056)

摘要:

睡眠障碍受到越来越多的关注,且自动化睡眠分期的准确性、泛化性受到了越来越多的挑战。然而,公开的睡眠数据十分有限,睡眠分期任务实际上更近似于一种小样本场景;同时由于睡眠特征的个体差异普遍存在,现有的机器学习模型很难保证准确判读未参与训练的新受试者的数据。为了实现对新受试者睡眠数据的精准分期,现有研究通常需要额外采集、标注新受试者的大量数据,并对模型进行个性化微调。基于此,借鉴迁移学习中基于缩放-偏移的权重迁移思想,提出一种元迁移睡眠分期模型MTSL(Meta Transfer Sleep Learner),设计了一种新的元迁移学习框架:训练阶段包括预训练与元迁移训练两步,其中元迁移训练时使用大量的元任务进行训练;而在测试阶段仅使用极少的新受试者数据进行微调,模型就能轻松适应新受试者的特征分布,大幅减少对新受试者进行准确睡眠分期的成本。在两个公开的睡眠数据集上的实验结果表明,MTSL模型在单数据集、跨数据集两种条件下都能取得更高的准确率和F1分数,这表明MTSL更适合小样本场景下的睡眠分期任务。

关键词: 睡眠分期, 小样本, 元学习, 迁移学习, 深度学习, 脑电信号

Abstract:

Sleep disorders are receiving more and more attention, and the accuracy and generalization of automated sleep stage classification are facing more and more challenges. However, due to the very limited human sleep data publicly available, the sleep stage classification task is actually similar to a few-shot scenario. And due to the widespread individual differences in sleep features, it is difficult for existing machine learning models to guarantee accurate classification of data from new subjects who have not participated in the training. In order to achieve accurate stage classification of new subjects’ sleep data, existing studies usually require additional collection and labeling of large amounts of data from new subjects and personalized fine-tuning of the model. Based on this, a new sleep stage classification model, Meta Transfer Sleep Learner (MTSL), was proposed. Inspired by the idea of Scale & Shift based weight transfer strategy in transfer learning, a new meta transfer learning framework was designed. The training phase included two steps: pre-training and meta transfer training, and many meta-tasks were used for meta transfer training. In the test phase, the model could be easily adapted to the feature distribution of new subjects by fine-tuning with only a few new subjects’ data, which greatly reduced the cost of accurate sleep stage classification for new subjects. Experimental results on two public sleep datasets show that MTSL model can achieve higher accuracy and F1-score under both single-dataset and cross-dataset conditions. This indicates that MTSL is more suitable for sleep stage classification tasks in few-shot scenarios.

Key words: sleep stage classification, few-shot, meta learning, transfer learning, deep learning, ElectroEncephaloGraphy (EEG)

中图分类号: