label shift
Description: A distribution shift in which class proportions change between training and test data, affecting machine-learning prediction and adaptation.
Primary topic: Domain Adaptation and Few-Shot Learning (44%)
OpenAlex ID: keywords/label-shift
