Tag
This paper presents a learning-based framework that models individual scorer behavior via confusion matrices to derive more robust sleep stage labels from multiple experts, improving accuracy on DOD-H and DOD-O datasets.