Machine Learning and Algorithms

Systematic comparison of incomplete-supervision approaches for biomedical imaging classification

The paper compares incomplete-supervision methods for biomedical image classification, finding that pre-training and semi-supervised learning contribute more than active learning to performance.