Skip to content
AI.info

AI.info

Kinds of learning

Supervised, unsupervised, self-supervised: the feedback signal decides everything downstream, including what can go wrong.

  1. Learning Approaches: Start With the Feedback
  2. Supervised Learning: Learning From Labeled Examples
  3. Labels, Targets, and the Limits of Ground Truth
  4. Classification: Categories, Scores, and Decisions
  5. Regression: Predicting Quantities Without Pretending to Know the Future
  6. Ranking, Scoring, and Structured Prediction
  7. Learning From Preferences and Pairwise Comparisons
  8. Unsupervised Learning: Searching for Structure Without Task Labels
  9. Clustering and Segmentation: Useful Groups Without Invented Essences
  10. Dimensionality Reduction and Representation Discovery
  11. Anomaly, Novelty, and Density: Finding What Does Not Fit
  12. Generative and Discriminative Modeling
  13. Reinforcement Learning: Learning From Actions and Consequences
  14. States, Actions, Rewards, and Policies
  15. Exploration, Exploitation, and Contextual Bandits
  16. Imitation Learning: Learning Behavior From Demonstrations
  17. Semi-Supervised Learning: A Few Labels, Many Unlabeled Examples
  18. Self-Supervised Learning: Creating Training Signals From the Data Itself
  19. Weak Supervision: Combining Noisy Sources Instead of Waiting for Perfect Labels
  20. Active Learning and Human-in-the-Loop Systems
  21. Transfer Learning and Domain Adaptation
  22. Batch, Online, and Continual Learning
  23. Prediction, Causation, and Counterfactual Questions
  24. Multi-Task and Hybrid Learning Systems
  25. Choosing the Learning Strategy Under Real Constraints
  26. Capstone: Design a Learning System and Defend Every Signal