For the complete list of unsupervised learning algorithms available in scikit-learn, please see the documentation.

• Manifold learning
• Clustering
• Decomposition
• Outlier detection
• Hidden markov models

Manifold learning for dimensionality reduction¶

Goal¶

The goal in manifold learning is to reduce the dimensionality of the dataset while still preserving its structure, so that similar samples are close together.

Applications¶

• Visualize sample space
• Determine whether two new samples are similar

Types¶

• Linear projections
• Locally linear embedding and its regularized variations

Practical considerations¶

• Scale samples
• Remove noisy data
• Remove identical samples

Exercise¶

Make a subset of the digits dataset using images from only three digits, then try two different manifold learning techniques using plot_lle_digits.py for reference.

Clustering¶

Decomposition¶

Outlier detection¶

Hidden markov models¶