Unsupervised Clustering of Individuals Sharing Selective Attentional Focus Using Physiological Synchrony

Research on brain signals as indicators of a certain attentional state is moving from laboratory environments to everyday settings. Uncovering the focus individuals in such settings challenging because there usually limited information about real-world events, well lack data context at hand that correctly labeled with respect individuals' state. In most approaches, needed train attention monitoring models. We here investigate whether unsupervised clustering can be combined physiological synchrony electroencephalogram (EEG), electrodermal activity (EDA), and heart rate automatically identify groups sharing without using knowledge sensory stimuli or any individuals. used an experiment which 26 participants listened audiobook interspersed emotional sounds beeps. Thirteen were instructed narrative 13 broad range commonly applied dimensionality reduction ordination techniques—further referred mappings—in combination algorithms two based synchrony. Analyses performed three modalities EEG, EDA, separately, all possible combinations these modalities. The best unimodal results obtained when applying yielding maximum accuracy 85%. Even though use EDA by itself did not lead accuracies significantly higher than chance level, combining EEG measures multimodal approach generally resulted classification only EEG. Additionally, found more consistent across data, making algorithm choice less important. Our finding into important support studies engagement