User guide: contents

• 1. Installing scikits.learn
  • 1.1. Installing an official release
    • 1.1.1. Installing from source
      • 1.1.1.1. Easy install
      • 1.1.1.2. From source package
    • 1.1.2. Windows installer
    • 1.1.3. Building on windows
  • 1.2. Third party distributions of scikits.learn
    • 1.2.1. Debian and derivatives (Ubuntu)
    • 1.2.2. Python(x, y)
    • 1.2.3. Enthought python distribution
    • 1.2.4. Macports
    • 1.2.5. NetBSD
  • 1.3. Bleeding Edge
  • 1.4. Testing
• 2. Getting started: an introduction to machine learning with scikits.learn
  • 2.1. Machine learning: the problem setting
  • 2.2. Loading an example dataset
  • 2.3. Learning and Predicting
  • 2.4. Model persistence
• 3. Supervised learning
  • 3.1. Generalized Linear Models
    • 3.1.1. Ordinary Least Squares (OLS)
      • 3.1.1.1. OLS Complexity
    • 3.1.2. Ridge Regression
      • 3.1.2.1. Ridge Complexity
      • 3.1.2.2. Generalized Cross-Validation
    • 3.1.3. Lasso
    • 3.1.4. Elastic Net
    • 3.1.5. Least Angle Regression
    • 3.1.6. LARS Lasso
      • 3.1.6.1. Mathematical formulation
    • 3.1.7. Bayesian Regression
      • 3.1.7.1. Bayesian Ridge Regression
    • 3.1.8. Automatic Relevance Determination - ARD
      • 3.1.8.1. Mathematical formulation
    • 3.1.9. Logisitic regression
    • 3.1.10. Stochastic Gradient Descent - SGD
  • 3.2. Support Vector Machines
    • 3.2.1. Classification
      • 3.2.1.1. Multi-class classification
      • 3.2.1.2. Unbalanced problems
    • 3.2.2. Regression
    • 3.2.3. Density estimation, outliers detection
    • 3.2.4. Support Vector machines for sparse data
    • 3.2.5. Complexity
    • 3.2.6. Tips on Practical Use
    • 3.2.7. Kernel functions
      • 3.2.7.1. Custom Kernels
        • 3.2.7.1.1. Using python functions as kernels
        • 3.2.7.1.2. Using the Gram matrix
    • 3.2.8. Mathematical formulation
      • 3.2.8.1. SVC
      • 3.2.8.2. NuSVC
    • 3.2.9. Implementation details
  • 3.3. Stochastic Gradient Descent
    • 3.3.1. Classification
    • 3.3.2. Regression
    • 3.3.3. Stochastic Gradient Descent for sparse data
    • 3.3.4. Complexity
    • 3.3.5. Tips on Practical Use
    • 3.3.6. Mathematical formulation
      • 3.3.6.1. SGD
    • 3.3.7. Implementation details
  • 3.4. Nearest Neighbors
    • 3.4.1. Classification
    • 3.4.2. Regression
    • 3.4.3. Efficient implementation: the ball tree
  • 3.5. Feature selection
    • 3.5.1. Univariate feature selection
      • 3.5.1.1. Feature scoring functions
        • 3.5.1.1.1. For classification
        • 3.5.1.1.2. For regression
  • 3.6. Gaussian Processes
    • 3.6.1. An introductory regression example
    • 3.6.2. Mathematical formulation
      • 3.6.2.1. The initial assumption
      • 3.6.2.2. The best linear unbiased prediction (BLUP)
      • 3.6.2.3. The empirical best linear unbiased predictor (EBLUP)
    • 3.6.3. Correlation Models
    • 3.6.4. Regression Models
    • 3.6.5. Implementation details
  • 3.7. Partial Least Squares
  • 3.8. Naive Bayes
    • 3.8.1. Gaussian Naive Bayes
• 4. Unsupervised learning
  • 4.1. Gaussian mixture models
    • 4.1.1. GMM classifier
  • 4.2. Clustering
    • 4.2.1. K-means
    • 4.2.2. Affinity propagation
    • 4.2.3. Mean Shift
    • 4.2.4. Spectral clustering
    • 4.2.5. Hierarchical clustering
      • 4.2.5.1. Adding connectivity constraints
  • 4.3. Decomposing signals in components (matrix factorization problems)
    • 4.3.1. Principal component analysis (PCA)
      • 4.3.1.1. Exact PCA and probabilistic interpretation
      • 4.3.1.2. Approximate PCA
      • 4.3.1.3. Kernel PCA
    • 4.3.2. Independent component analysis (ICA)
    • 4.3.3. Non-negative matrix factorization (NMF)
  • 4.4. Covariance estimation
    • 4.4.1. Empirical covariance
    • 4.4.2. Shrunk Covariance
      • 4.4.2.1. Basic shrinkage
      • 4.4.2.2. Ledoit-Wolf shrinkage
      • 4.4.2.3. Oracle Approximating Shrinkage
• 5. Model Selection
  • 5.1. Cross-Validation
    • 5.1.1. Examples
      • 5.1.1.1. Leave-One-Out - LOO
      • 5.1.1.2. Leave-P-Out - LPO
      • 5.1.1.3. K-fold
      • 5.1.1.4. Stratified K-Fold
      • 5.1.1.5. Leave-One-Label-Out - LOLO
      • 5.1.1.6. Leave-P-Label-Out
  • 5.2. Grid Search
    • 5.2.1. Examples
    • 5.2.2. Notes
• 6. Dataset loading utilities
  • 6.1. Datasets shipped with the scikit learn
  • 6.2. Dataset generators
  • 6.3. The Labeled Faces in the Wild face recognition dataset
    • 6.3.1. Usage
    • 6.3.2. Examples
  • 6.4. The 20 newsgroups text dataset
    • 6.4.1. Usage
    • 6.4.2. Examples
• 7. Class Reference
  • 7.1. Support Vector Machines
    • 7.1.1. For sparse data
    • 7.1.2. Low-level methods
  • 7.2. Generalized Linear Models
    • 7.2.1. For sparse data
  • 7.3. Naive Bayes
  • 7.4. Nearest Neighbors
  • 7.5. Gaussian Mixture Models
  • 7.6. Hidden Markov Models
  • 7.7. Clustering
  • 7.8. Metrics
    • 7.8.1. Pairwise metrics
  • 7.9. Covariance Estimators
    • 7.9.1. Covariance estimators
  • 7.10. Signal Decomposition
  • 7.11. Linear Discriminant Analysis
  • 7.12. Cross Validation
  • 7.13. Grid Search
  • 7.14. Feature Selection
  • 7.15. Feature Extraction
    • 7.15.1. From images
    • 7.15.2. From text
  • 7.16. Pipeline
  • 7.17. Partial Least Squares