Cross-Population Evaluation

Biomedical signal evaluation should preserve population identity and report directional cohort-level transfer. The n-gram design enumerated all source-population combinations and tested each remaining population as a held-out target. The study treated five independently acquired resting-state EEG cohorts as separate populations throughout evaluation.