학술논문

LATTE: Label-efficient incident phenotyping from longitudinal electronic health records