ConceptNon-Exhaustive Supervision
Non-Exhaustive Supervision
An annotation scenario in machine learning in which missing labels in training data do not represent negative samples but instead reflect incomplete annotation. This condition is common in domains such as medical imaging report generation.
Timeline (last 90 days)
Oct 5
Dental CBCT reports have a non-exhaustive supervision problem, where 'not mentioned' in a report can indicate either negative evidence or a missing annotation.
Unverified50%