1. PU Learning and its sources

Content

• What is PU Learning?
• 10 Sources of PU data
  1. Automatic diagnosis
  2. Positive examples are easier to obtain
  3. Indirect labels
  4. Case-Control
  5. Negative-class datashift
  6. Under-reporting
  7. One-class classification
  8. Inlier-based outlier detection
  9. Knowledge base completion
  10. Identification