A situation in machine learning where certain classes are underrepresented in the training data, potentially leading to biased model predictions.
Last updated 10 months ago