Introduction to Unsupervised Learning
- Unsupervised learning: learning without the desired output (‘teacher’ signals).
Some methods are:
- Dimensionality reduction (e.g. PCA).
- Association Rules/Recommender Systems.
- Clustering: one of the widely-used unsupervised learning methods.
More distance metrics:
- Correlation.
- Minkowski.
- Mahalanobis.
They are often application dependant. The important things are the shape, distance and scale.