[PYTHON] [Translation] scikit-learn 0.18 User Guide Table of Contents

Google translated http://scikit-learn.org/0.18/user_guide.html Tutorial here

User guide

1. Supervised learning

1.1. Generalized linear model : us: untranslated

1. Least squares
2. The complexity of the least squares method
3. Ridge regression
4. Ridge complexity
5. Normalization parameter setting: Generalized mutual validation
6. Minimum absolute contraction and selection operator --Lasso
7. Setting regularization parameters
8. Use of mutual verification
9. Information-based model selection
10. Multitasking Lasso
11. Elastic Net
12. Multitasking Elastic Net
13. Minimum Angle Regression-LARS
14. LARS Lasso
15. Mathematical prescription
16. Orthogonal Matching Pursuit (OMP)
17. Bayesian regression
18. Bayesian ridge regression
19. Automatic validation --ARD
20. Logistic regression
21. Stochastic Gradient Descent-SGD
22. Perceptron
23. Passive aggressive algorithm
24. Robustness regression: outliers and modeling errors
25. Different scenarios and useful concepts 2. RANSAC：RANdom SAmple Consensus
26. Algorithm details
27. Theil-Sen Estimator: Generalization-Median-based estimator
28. Theoretical consideration
29. Hoover regression
30. Notes
31. Polynomial regression: Extension of linear model with basis functions

1.2. Linear and quadratic discriminant analysis untranslated

1. Dimensionality reduction using linear discriminant analysis
2. Mathematical formulation of LDA and QDA classifiers
3. Mathematical formulation of LDA dimensionality reduction
4. Shrinkage
5. Estimating algorithm

1.3. Kernel Ridge Regression : us: Untranslated

1.4. Support Vector Machine: jp:

1. Classification
2. Multi-class classification
3. Score and probability
4. Imbalanced problem
5. Regression
6. Density estimation, novelty detection
7. Complex
8. Practical tips
9. Kernel function
10. Custom kernel
11. Use Python functions as kernel
12. Use of Gram matrix
13. RBF kernel parameters
14. Mathematical prescription
    1. SVC
    2. NuSVC
    3. SVR
15. Implementation details

1.5. Stochastic Gradient Descent : us: Untranslated

1. Classification
2. Regression
3. Stochastic gradient descent of sparse data
4. Complex
5. Practical tips
6. Mathematical prescription
   1. SGD
7. Implementation details

1.6. Nearest neighbor method : us: untranslated

1. Unsupervised nearest neighbor method
2. Find the closest neighbor
3. KDTree class and BallTree class
4. Nearest neighbor classification
5. Nearest neighbor regression
6. Nearest neighbor algorithm
7. Brute force
8. K-D tree
9. Ball tree
10. Selection of nearest neighbor algorithm
11. Effect of leaf_size
12. Nearest center of gravity classifier
13. Closest Shrinken Centroid
14. Approximate neighborhood nearby
15. Community sensitive hashing forest
16. Mathematical description of local sensitivity hash

1.7. Gaussian process : us: untranslated

1. Gaussian process regression (GPR)
2. GPR example
3. GPR with noise level estimation
4. Comparison of GPR and kernel ridge regression
5. GPR of Mauna Loa CO2 data
6. Gaussian process classification (GPC)
7. GPC example
8. Probabilistic prediction by GPC
9. Diagram of GPC on XOR dataset
10. Gaussian process classification (GPC) in the iris dataset
11. Gaussian process kernel
12. Gaussian process kernel API
13. Basic kernel
14. Kernel operator
15. Radial basis function (RBF) kernel
16. Matteran kernel
17. Reasonable secondary kernel
18. Exp-Sine-Squared kernel
19. Dot product kernel
20. References
21. Legacy Gaussian process
22. Example of introduction regression
23. Fitting noisy data
24. Mathematical prescription
25. First assumption
26. Best Linear Unbiased Predictions (BLUP)
27. Empirically Best Linear Bias Predictor (EBLUP)
28. Correlation model
29. Regression model
30. Implementation details

1.8. Cross decomposition : us: untranslated

1.9. Naive Bayes : us: Untranslated

1. Gauss Naive Bayes
2. Polynomial naive bayes
3. Bernoulli Naive Bayes
4. Out-of-core naive Bayes model fitting

1.10. Decision tree : us: untranslated

1. Classification
2. Regression
3. Multi-output problem
4. Complex
5. Practical tips
6. Tree algorithm: ID3, C4.5, C5.0 and CART
7. Mathematical prescription
8. Classification criteria
9. Regression criteria

1.11. Ensemble method: jp:

1. Bagging meta estimator
2. Randomized tree forest
3. Random forest
4. Very random tree
5. Parameters
6. Parallelization
7. Functional importance evaluation
8. Fully random tree embedding
9. AdaBoost
10. Usage
11. Gradient Tree Boost
12. Classification
13. Regression
14. Fit additional weak learners
15. Tree size control
16. Mathematical prescription
17. Loss function
18. Normalization
19. Shrinkage
20. Subsampling
21. Interpretation
22. Importance of function
23. Partially dependent
24. VotingClassifier
25. Most class labels (majority vote / carefully selected)
26. Usage
27. Weighted average probability (soft voting)
28. Use Voting Classifier with Grid Search
29. Usage

1.12. Multi-class algorithm and multi-label algorithm: jp:

1. Multi-label classification format
2. One rest
3. Multi-class learning
4. Multi-label learning
5. One-on-one
6. Multi-class learning
7. Error correction output code
8. Multi-class learning
9. Multi-output regression
10. Classification of multiple outputs

1.13. Feature selection: jp:

1. Delete features with low variance
2. Selection of univariate function
3. Recursive feature removal
4. Feature selection using SelectFromModel
5. L1-based feature selection
6. Random sparse model
7. Tree-based feature selection
8. Selection of functions as part of the pipeline

1.14. Semi-supervised : us: untranslated

1. Label propagation

1.15. Isotonic regression: jp:

1.16. Probability calibration: jp:

1.17. Neural network model (supervised) : us: untranslated

1. Multilayer perceptron
2. Classification
3. Regression
4. Normalization
5. Algorithm
6. Complex
7. Mathematical prescription
8. Practical tips
9. More control with warm_start

2. Unsupervised learning

2.1. Gaussian mixed model : us: untranslated

1. Gauss mixed
2. Advantages and disadvantages of Gaussian Mixture
3. Advantages
4. Disadvantages 2 Selection of the number of components in a classical Gaussian mixed model
5. Estimate algorithm maximize expected value
6. Variational Bayes Gauss mixed
7. Estimating algorithm: variational inference
8. Advantages and disadvantages of transformation reasoning with BayesianGaussianMixture
9. Advantages
10. Disadvantages
11. Dirichlet process

2.2. Manifold learning : us: untranslated

1. Introduction
2. Isomap
3. Complex
4. Locally linear embedding
5. Complex
6. Locally modified linear embedding
7. Complex
8. Unique mapping of Hessian matrix
9. Complex
10. Spectrum embedding
11. Complex
12. Local tangent space alignment
13. Complex
14. Multidimensional scaling (MDS)
15. Metric MDS
16. Non-metric MDS
17. t-Distributed Stochastic Neighborhood Embedding (t-SNE)
18. t-SNE optimization
19. Burns hat t-SNE
20. Practical tips

2.3. Clustering : us: Untranslated

1. Overview of clustering method
2. K-means
3. Mini batch K-Means
4. Affinity propagation
5. Average shift
6. Spectrum clustering
7. Difference in label assignment method
8. Hierarchical clustering
9. Different linkage types: Ward, full average linkage
10. Add connection constraints
11. Change metric
12. Density-based spatial clustering (DBSCAN)
13. Hierarchical balanced iterative reduction and clustering (BIRCH)
14. Clustering performance evaluation
15. Adjusted land index
16. Benefits
17. Disadvantages
18. Mathematical prescription
19. Mutual information-based scoring
20. Benefits
21. Disadvantages
22. Mathematical prescription
23. Homogeneity, integrity and V-scale
24. Benefits
25. Disadvantages
26. Mathematical prescription
27. Fowlkes-Mallows score
28. Benefits
29. Disadvantages
30. Silhouette coefficient
31. Benefits
32. Disadvantages
33. Karinsky Harabaz Index
34. Benefits
35. Disadvantages

2.4. Biclustering : us: Untranslated

1. Spectral co-clustering
2. Mathematical prescription
3. Spectrum by clustering
4. Mathematical prescription
5. Bi-clustering evaluation

2.5. Decompose the signal in the component (matrix factorization problem): jp:

1. Principal component analysis (PCA)
2. Accurate PCA and stochastic interpretation
3. Incremental PCA
4. PCA with randomized SVD
5. Kernel PCA
6. Sparse PCA and MiniBatchSparsePCA
7. Truncation singular value decomposition and latent semantic analysis
8. Dictionary learning
9. Sparse coding with a pre-computed dictionary
10. General dictionary learning
11. Mini-batch dictionary learning
12. Factor analysis
13. Independent Component Analysis (ICA)
14. Non-negative matrix factorization (NMF or NNMF)
15. Latent Dirichlet Allocation (LDA)

2.6. Covariance estimation : us: untranslated

1. Empirical covariance
2. Reduced covariance
3. Basic contraction
4. Ledoit-Wolf Shrink
5. Oracle approximate contraction
6. Sparse inverse covariance
7. Robust covariance estimation
8. Minimum covariance determinant

2.7. Detection of novelty and outliers: jp:

1. Detection of novelty
2. Outlier detection
3. Install the oval envelope
4. Isolation forest 3.1 Class SVM vs. Elliptical Envelope vs. Isolation Forest

2.8. Density estimation: jp:

1. Density estimation: Histogram
2. Kernel density estimation

2.9. Neural network model (unsupervised) : us: untranslated

1. Limited Boltzmann machine
2. Graphical model and parameterization
3. Bernoulli restricted Boltzmann machine
4. Stochastic maximum likelihood learning

3. Model selection and evaluation

3.1. Cross-validation: Evaluate estimator performance: jp:

1. Calculation of cross-validated metrics
2. Obtaining forecasts by cross-validation
3. Cross-validation iterator
4. i.i.d cross-validation iterator data
5. K times 2. Leave One Out（LOO） 3. Leave P Out（LPO）
6. Random Permutation Mutual Verification a.k.a. Shuffle & Split
7. Mutual validation iterator with hierarchy based on class label
8. Layered K times
9. Layered shuffle split
10. Mutual validation iterator for grouped data
11. Group k times
12. Leave one group
13. Leave P group
14. Group shuffle split
15. Predefined Fold-Splits / Validation-Sets
16. Mutual verification of time series data
17. Time series division
18. Shuffle precautions
19. Mutual validation and model selection

3.2. Tuning of hyperparameters of estimator: jp:

1. Complete grid search
2. Randomized parameter optimization
3. Parameter search tips
4. Objective metric specification
5. Composite estimates and parameter space
6. Model selection: development and evaluation
7. Parallel
8. Robustness to disability
9. Alternative to brute force parameter search
10. Model-specific mutual validation
11. Information standards
12. Other estimators

3.3. Model evaluation: Quantify the quality of prediction: jp:

1. Score parameter: Definition of model evaluation rule
2. Common case: predefined values
3. Define a scoring strategy from a metric function
4. Implementation of your own scoring object
5. Classification metric
6. From binary to multi-class, multi-label
7. Accuracy score
8. Cohen's kappa
9. Confusion Matrix
10. Classification report
11. Humming loss
12. Jacquard similarity coefficient score
13. Precision, recall, F-measures
14. Binary classification
15. Multi-class and multi-label classification
16. Hinge loss
17. Log loss
18. Matthews correlation coefficient
19. Receiver Operating Characteristic (ROC)
20. Zero one loss
21. Breather score loss
22. Multi-label ranking metric
23. Coverage error
24. Average accuracy of label rank
25. Loss of ranking
26. Regression metric
27. Explained variance score
28. Average absolute error
29. Mean squared error
30. Central absolute error
31. R² score, coefficient of determination
32. Clustering metric
33. Dummy estimator

3.4. Model persistence: flag_jp:

1. Persistence example
2. Security and maintainability limits

3.5. Verification curve: Plot the score to evaluate the model: jp:

1. Verification curve
2. Learning curve

4. Data set conversion

4.1. Pipeline and Feature Union: Estimator combination: jp:

1. Pipeline: Chain estimator
2. Usage
3. Note
4. FeatureUnion: Composite feature space
5. Usage

4.2. Feature extraction: jp:

1. Loading features from dicts
2. Feature hash
3. Implementation details
4. Text feature extraction
5. Word notation
6. Rarity
7. How to use the common vectorizer
8. Weighting of Tf-idf terms
9. Decoding text files
10. Applications and samples
11. Limitations of expression in Bag of Words
12. Vectorize a large text corpus using hash tricks
13. Perform out-of-core scaling with HashingVectorizer
14. Customized vectorizer class
15. Image feature extraction
16. Patch extraction
17. Image connectivity graph

4.3. Data preprocessing: jp:

1. Standardization, averaging and variance scaling
2. Scaling features to range
3. Sparse data scaling
4. Scaling data containing outliers
5. Centering kernel matrix
6. Normalization
7. Binarization
8. Feature binarization
9. Encode the function of the category
10. Completion of missing values
11. Generate polynomial features
12. Custom transformer

4.4. Unsupervised dimensionality reduction: jp:

1. PCA: Principal component analysis
2. Random projection
3. Feature agglomeration

4.5. Random projection: jp:

1. Johnson-Lindenstrauss lemma
2. Gauss random projection
3. Sparse random projection

4.6. Kernel approximation : us: untranslated

1. Nystroem method for kernel approximation
2. Radial basis function kernel
3. Additive Chi Squared Kernel
4. Skewed chi-square kernel
5. Math details

4.7. Pairwise Metrics, Similarities and Kernels

1. Cosine similarity
2. Linear kernel
3. Polynomial kernel
4. Sigmoid kernel
5. RBF kernel
6. Laplacian kernel
7. Chi-square kernel

4.8. Transform the prediction target (y): jp:

1. Label binarization
2. Label encoding

5. Dataset reading utility : us: untranslated

1. General dataset API
2. Toy dataset
3. Sample image
4. Sample generator
5. Generator for classification and clustering
6. Single label
7. Multi-label 3. Biclustering
8. Regression generator
9. Generator for diverse learning
10. Disassembly generator
11. Dataset in svmlight / libsvm format
12. Loading from an external dataset
13. Olivetti faces dataset
14. 20 newsgroup text datasets
15. Usage
16. Convert text to vector
17. Text filtering for more realistic training
18. Download the dataset from the mldata.org repository
19. Labeled faces in the wild face recognition dataset
20. Usage
21. Example
22. Deforestation
23. RCV1 dataset
24. Boston Home Price Dataset
25. Note
26. Breast Cancer Wisconsin (Diagnosis) Database
27. Note
28. References
29. Diabetes dataset
30. Note
31. Optical recognition of handwritten digit data
32. Note
33. References
34. Iris plant database
35. Note
36. References
37. Linnerrud dataset
38. Note
39. References

6. Computational Expansion Strategy: Larger Data : us: Untranslated

1. Scaling instances using out-of-core learning
2. Streaming instance
3. Feature extraction
4. Incremental learning
5. Example
6. Notes

7. Computation performance : us: untranslated

1. Predicted latency
2. Bulk vs. atomic mode
3. Impact of number of features
4. Impact of input data representation
5. Impact of model complexity
6. Feature extraction latency
7. Predicted throughput
8. Tips and techniques
9. Linear algebra library
10. Model compression
11. Model shape change
12. Link

Tutorial here

© 2010 --2016, scikit-learn developers (BSD license).