An enhanced ensemble deep random vector functional link network for driver fatigue recognition

• DOI: 10.1016/j.engappai.2023.106237
• Publication Date: 2023-04-08T03:45:01Z
• Authors: Ruilin Li; Ruobin Gao; Liqiang Yuan; P.N. Suganthan; Lipo Wang; Olga Sourina
• Keywords: Extractor; Feature (linguistics); Ensemble Learning; Feature vector

This work investigated the use of an ensemble deep random vector functional link (edRVFL) network for electroencephalogram (EEG)-based driver fatigue recognition. Against low feature learning capability edRVFL from raw EEG signals, two strategies were exploited in this work. Specifically, first one was to exploit advantages extractor module CNNs, i.e., CNN features as input network. The second improve An enhanced edRFVL named FGloWD-edRVFL proposed, which four enhancements implemented, including forest-based Feature selection, Global output layer, Weighting and entropy-based Dynamic ensemble. proposed evaluated on challenging cross-subject recognition tasks. results indicated that model could boost performance, significantly outperforming all strong baselines. step-wise analysis further demonstrated effectiveness