Focuses on the extraction and fusion of neural features such as event-related potentials (ERPs) and phase–amplitude coupling (PAC), and on characterizing domain shift and developing adaptive generalization methods for EEG decoding across subjects and devices, with explainable and deployable applications in sleep health and cognitive assessment. It also develops adaptive discrimination methods for motor-imagery brain–computer interfaces (BCIs) combining machine learning and domain adaptation, and investigates the cognitive mechanisms of auditory ERPs using auditory oddball paradigms, including enhanced responses to self-name stimuli and asymmetric response mechanisms of auditory processing, with implications for assessing neuropsychiatric conditions.