- Unsupervised Learning: No output (Y) is given; Try to learn structure (X).
- Supervised Learning: Both input (X), and output (Y) are given; Try to predict label / real-value.
- Classification: Try to predict label.
- Regression: Try to predict real-value.