The real-time assessment of mental workload (MWL) is critical for development of intelligent human–machine cooperative systems in various safety–critical applications. Although data-driven machine learning (ML) approach has shown promise in MWL recognition, there is still difficulty in acquiring a sufficient number of labeled data to train the ML models. This paper proposes a semi-supervised extreme learning machine (SS-ELM) algorithm for MWL pattern classification requiring only a small number of labeled data. The measured data analysis results show that the proposed SS-ELM paradigm can effectively improve the accuracy and efficiency of MWL classification and thus provide a competitive ML approach to utilizing a large number of unlabeled data which are available in many real-world applications.

精神工作量（MWL）的实时评估对于在各种安全关键型应用中开发智能人机协作系统至关重要。尽管数据驱动的机器学习（ML）方法已在MWL识别中显示出希望，但仍难以获取足够数量的标记数据来训练ML模型。本文提出了一种半监督的极限学习机（SS-ELM）算法，用于只需要少量标记数据的MWL模式分类。实测数据分析结果表明，提出的SS-ELM范式可以有效地提高MWL分类的准确性和效率，从而提供一种竞争性的ML方法，以利用在许多实际应用中可用的大量未标记数据。