In this tutorial, you will learn everything that you need to know about the Scikit-Learn Machine Learning library.

We willc over all the scikit-learn modules and method available, each tutorial including Python examples.

Navigate to the module specific tutorial to find out how to use it in Python.

Different Scikit-Learn Modules

Here is a full list of the Scikit-learn available modules.

• sklearn.calibration: Used for probability calibration.
• sklearn.cluster: Used for clustering data.
• sklearn.compose: Used for creating composite estimators.
• sklearn.covariance: Used for covariance estimation.
• sklearn.cross_decomposition: Used for cross decomposition.
• sklearn.datasets: Used for loading and generating datasets.
• sklearn.decomposition: Used for matrix decomposition.
• sklearn.discriminant_analysis: Used for discriminant analysis.
• sklearn.dummy: Used for creating dummy estimators.
• sklearn.ensemble: Used for ensemble methods.
• sklearn.exceptions: Used for handling exceptions and warnings.
• sklearn.experimental: Used for experimental features.
• sklearn.feature_extraction: Used for feature extraction.
• sklearn.feature_selection: Used for feature selection.
• sklearn.gaussian_process: Used for Gaussian processes and kernels.
• sklearn.impute: Used for imputation.
• sklearn.inspection: Used for model inspection and plotting.
• sklearn.isotonic: Used for isotonic regression.
• sklearn.kernel_approximation: Used for kernel approximation.
• sklearn.kernel_ridge: Used for kernel ridge regression.
• sklearn.linear_model: Used for linear models and classifiers.
• sklearn.manifold: Used for manifold learning.
• sklearn.metrics: Used for model evaluation metrics.
• sklearn.mixture: Used for Gaussian mixture models.
• sklearn.model_selection: Used for model selection and validation.
• sklearn.multiclass: Used for multiclass classification.
• sklearn.multioutput: Used for multioutput regression and classification.
• sklearn.naive_bayes: Used for Naive Bayes classifiers.
• sklearn.neighbors: Used for nearest neighbors algorithms.
• sklearn.neural_network: Used for neural network models.
• sklearn.pipeline: Used for constructing pipelines.
• sklearn.preprocessing: Used for data preprocessing and normalization.
• sklearn.random_projection: Used for random projection.
• sklearn.semi_supervised: Used for semi-supervised learning.
• sklearn.svm: Used for support vector machines.
• sklearn.tree: Used for decision trees and plotting.
• sklearn.utils: Used for various utility functions and classes.

Foundation Classes and Utilities

In this section, you will learn the foundational aspects of scikit-learn with our base classes and utility functions (sklearn.base). These essential building blocks serve as the backbone for constructing powerful machine learning models and conducting insightful analyses, ensuring robustness and versatility in your projects.

The sklearn.base library is for base classes and utilities.

Data Preprocessing and Normalization Functions

Use Data preprocessing and normalization to build robust machine learning models. The sklearn.preprocessing module offers a lot of tools for data preprocessing and normalization. Whether you’re dealing with missing values, categorical variables, or numerical features, this module provides a wide range of techniques for transforming and standardizing your data prior to model training. From scaling and encoding to imputation and feature engineering, sklearn.preprocessing offers flexible solutions for preparing your data and optimizing model performance across diverse machine learning tasks and domains.

Clustering Algorithms

The sklearn.cluster library is used for clustering classes and functions.

Learn about various clustering algorithms with sklearn.cluster, featuring a comprehensive collection of classes and functions for partitioning data into cohesive groups. Whether you’re exploring K-means, DBSCAN, or hierarchical clustering, this module provides the tools you need to uncover meaningful patterns in your data.

Covariance Estimation Methods

Explore covariance estimation techniques with sklearn.covariance, where you can analyze the relationships between variables and uncover underlying patterns in your data. This module offers a range of estimators for robust covariance estimation, ensuring accurate and reliable results in your analyses.

Cross Decomposition Techniques

The sklearn.cross_decomposition library is for cross decomposition.

Cross decomposition methods with sklearn.cross_decomposition offers powerful techniques for modeling relationships between multiple datasets. Whether you’re performing canonical correlation analysis or partial least squares regression, this module provides the tools you need to uncover latent structures and extract valuable insights from your data.

Built-in Datasets and Data Loaders

Scikit-learn offers various datasets with their sklearn.datasets module, a collection of sample data for experimentation and analysis. From classic benchmark datasets to custom-generated samples, this module offers a rich repository of data to support your machine learning endeavors.

The sklearn.datasets library is for dataset loaders and generators.

Matrix Decomposition Algorithms

Matrix decomposition algorithms are accessible with sklearn.decomposition. From principal component analysis (PCA) to singular value decomposition (SVD), this module offers a range of techniques to reduce dimensionality and uncover latent patterns in your data, empowering you to make informed decisions and gain deeper insights into your datasets.

Discriminant Analysis Methods

Sklearn.discriminant_analysis, offers discriminant analysis methods for classification and dimensionality reduction. Whether you’re exploring linear discriminant analysis (LDA) or quadratic discriminant analysis (QDA), this module provides the means to identify discriminative features and build robust classifiers for your machine learning tasks.

Dummy Estimators for Baseline Performance

Explore the fundamentals of machine learning with sklearn.dummy, where you can build baseline models and assess the performance of more sophisticated algorithms. Whether you’re evaluating classification accuracy or regression performance, this module offers simple yet effective estimators to serve as benchmarks for your analyses, enabling you to gauge the efficacy of your models and refine your approach accordingly.

Ensemble Learning Methods

Ensemble methods are powerful in machine Learning. Scikit-learn ensemble methods are found in sklearn.ensemble module. From random forests to gradient boosting, this module offers a rich array of algorithms to ensemble learners and amplify the predictive power of your machine learning models, ensuring robustness and reliability in your predictions.

Feature Extraction Techniques

Unlock the latent information embedded in your data with sklearn.feature_extraction, offering powerful tools for extracting meaningful features from images and text. Whether you’re analyzing image datasets or processing textual documents, this module provides a versatile toolkit for feature extraction, enabling you to uncover valuable insights and drive actionable decisions from your data.

Feature Selection Algorithms

Feature Selection algorithms in Scikit-learn allow you to streamline your machine learning pipelines with sklearn.feature_selection. This module allows you to select optimal feature subsets and enhance model performance. Whether you’re reducing dimensionality or improving interpretability, this module offers a suite of techniques for feature selection, empowering you to focus on the most relevant attributes and boost the efficiency of your models.

Evaluation Metrics and Performance Measures

Navigate the landscape of model evaluation and performance metrics with sklearn.metrics, a comprehensive module offering a wide range of evaluation measures for classification, regression, clustering, and beyond. From accuracy and precision to silhouette score and adjusted Rand index, this module provides an extensive collection of metrics for assessing the quality and robustness of your machine learning models. Whether you’re fine-tuning hyperparameters or comparing different algorithms, sklearn.metrics equips you with the tools you need to make informed decisions and optimize your models for maximum performance.

Gaussian Process Models and Kernels

Gaussian Processes and Kernels with sklearn.gaussian_process. Explore powerful techniques for probabilistic regression and classification. From radial basis function kernels to Matérn covariance functions, this module offers a rich variety of tools for modeling complex relationships in your data and making robust predictions with uncertainty estimates, empowering you to tackle a wide range of machine learning tasks with confidence and accuracy.

Data Imputation Strategies

Use sklearn.impute to navigate missing data in Python. This module allows for imputing missing values and ensuring the integrity of your datasets. Whether you’re dealing with incomplete observations or sparse matrices, this module provides a range of strategies, from mean imputation to iterative imputation, to handle missing data effectively and prevent bias in your analyses, enabling you to extract reliable insights and make informed decisions from your data.

Probability Calibration Techniques

Let’s learn probability calibration with sklearn.calibration, where you can refine the calibration of your classifiers and enhance the reliability of your predictions. This module offers a suite of tools to fine-tune probability estimates and optimize the performance of your models.

Composite Estimators and Pipelines

Sklearn.compose is module for constructing composite estimators and seamlessly integrating preprocessing steps into your pipeline. By combining transformers and estimators, you can streamline your analysis and achieve optimal performance with ease.

Model Inspection and Plotting Utilities

The sklearn.inspection is used for inspection and plotting, model interpretation and visualization. Whether you’re exploring feature importances or dissecting model behavior, this module provides a comprehensive suite of techniques for inspecting and understanding your models, empowering you to diagnose issues, identify strengths, and fine-tune your algorithms for optimal performance.

Isotonic Regression Methods

Scikit-learn isotonic methods are used for non-parametric regression for monotonic relationships. Sklearn.isotonic is a module for fitting monotonic functions to your data and capturing non-linear relationships. Whether you’re modeling dose-response curves or calibrating probability estimates, this module offers robust algorithms for isotonic regression, enabling you to uncover hidden patterns and make accurate predictions in diverse machine learning applications.

Kernel Approximation Techniques

Perform kernel approximation with sklearn.kernel_approximation. If offers methods to approximate kernel functions and scale kernel-based algorithms to large-scale datasets. Whether you’re working with non-linear data or high-dimensional feature spaces, this module provides a range of techniques, from random Fourier features to Nystroem approximation, to approximate kernel matrices and accelerate computation, enabling you to tackle complex machine learning tasks with ease and efficiency.

Kernel Ridge Regression Algorithms

Kernel Ridge Regression with sklearn.kernel_ridge. This module offers algorithms for fitting non-linear models to your data and capturing complex relationships. Whether you’re modeling time series data or forecasting future trends, this module provides robust techniques for kernel-based regression, enabling you to make accurate predictions and extract valuable insights from your datasets.

Linear Models and Regressors

Get access to Scikit-learn linear models with sklearn.linear_model. This is very useful for regression and classification tasks. From traditional linear regressors to Bayesian models, this module offers a diverse range of algorithms for modeling linear relationships and making predictions with ease. Whether you’re dealing with high-dimensional data or noisy observations, sklearn.linear_model provides robust solutions, including outlier-robust regressors and generalized linear models, empowering you to extract valuable insights and uncover hidden patterns in your datasets.

Manifold Learning Algorithms

Manifold learning algorithms are techniques used in machine learning to uncover the underlying structure of high-dimensional data by reducing its dimensionality while preserving essential geometric properties. Scikit-learn offers those with the sklearn.manifold module. From Isomap to t-SNE, this module offers a rich array of algorithms for uncovering the underlying structure of your data and revealing intricate relationships between data points. Get deeper insights and enhancing your understanding of the underlying data manifold

Gaussian Mixture Models

Immerse yourself in the world of Gaussian Mixture Models with sklearn.mixture, offering powerful algorithms for modeling complex data distributions and uncovering hidden patterns. Whether you’re clustering high-dimensional data or performing density estimation, this module provides robust techniques for fitting mixture models to your data and extracting meaningful insights. With support for various covariance types and initialization methods, sklearn.mixture empowers you to tackle diverse machine learning tasks and discover latent structures within your datasets.

Model Selection and Validation Tools

Model selection and validation is paramount in machine learning and can be done with sklearn.model_selection. It offers tools for splitting datasets, optimizing hyperparameters, and assessing model performance. From cross-validation strategies to grid search and randomized search, this module provides flexible techniques for fine-tuning your models and preventing overfitting. Whether you’re building predictive models or evaluating classification algorithms, sklearn.model_selection equips you with the resources you need to make informed decisions and build robust machine learning pipelines.

Naive Bayes Classifiers

Naive Bayes classifiers are within the sklearn.naive_bayes module. This one offers probabilistic algorithms for classification tasks. Despite its naive assumption of feature independence, Naive Bayes remains a popular choice for its efficiency and ease of implementation, making it particularly suitable for large-scale datasets with high-dimensional features. Whether you’re classifying text documents or analyzing biological data, sklearn.naive_bayes provides reliable solutions for a wide range of classification problems, empowering you to build accurate and efficient predictive models with minimal computational overhead.

Nearest Neighbors Algorithms

Nearest neighbors algorithms are found within the sklearn.neighbors module, offering both supervised and unsupervised learning capabilities. Whether you’re performing classification, regression, or density estimation, this module provides intuitive and flexible techniques for making predictions based on the similarity of data points. From k-nearest neighbors to kernel density estimation, sklearn.neighbors offers a range of algorithms suited to different problem domains and data distributions, empowering you to build robust and accurate models for a variety of machine learning tasks.

Multiclass Classification Strategies

Perform multiclass classification with sklearn.multiclass. This will allow you to handle multi-label and multi-output problems. From one-vs-one to one-vs-rest approaches, this module provides a range of techniques for extending binary classifiers to multiclass scenarios and making accurate predictions across multiple classes. Whether you’re classifying text documents or analyzing images, sklearn.multiclass equips you with the tools you need to build versatile and scalable classification models capable of handling diverse datasets and complex tasks.

Multioutput Regression and Classification Techniques

Use sklearn.multioutput module for multioutput regression and classification when you have multiple target variables. Whether you’re predicting several continuous or categorical outputs simultaneously, this module offers a range of algorithms and techniques tailored to your needs. From extending traditional single-output models to handling correlated outputs, sklearn.multioutput provides robust solutions for addressing complex prediction tasks and extracting valuable insights from multi-dimensional data.

Neural Network Models

Dive into the realm of neural network models with sklearn.neural_network, a powerful module offering algorithms for training and deploying artificial neural networks. Whether you’re building feedforward networks for classification tasks or recurrent networks for sequence prediction, this module provides a rich set of tools for designing and fine-tuning neural architectures to suit your specific needs. With support for various activation functions, optimization algorithms, and regularization techniques, sklearn.neural_network equips you with the flexibility and scalability needed to tackle complex machine learning problems and achieve state-of-the-art performance.

Data Processing Pipelines

Streamline your machine learning workflow with sklearn.pipeline, a versatile module offering tools for constructing and managing data processing pipelines. Whether you’re performing feature scaling, feature selection, or model training, this module provides a unified interface for chaining together multiple transformers and estimators into a single pipeline. From data preprocessing to model evaluation, sklearn.pipeline simplifies the development and deployment of machine learning workflows, enabling you to iterate quickly and efficiently while maintaining reproducibility and scalability.

Random Projection Methods

Explore the power of random projection techniques with sklearn.random_projection, a module offering algorithms for dimensionality reduction and data compression. Whether you’re dealing with high-dimensional data or seeking to reduce computational complexity, this module provides efficient and scalable methods for projecting your data onto lower-dimensional subspaces while preserving key structural properties. From Gaussian random projection to sparse random projection, sklearn.random_projection offers a range of techniques for accelerating computation and improving the efficiency of machine learning algorithms across diverse applications and domains.

Semi-Supervised Learning Algorithms

Harness the power of semi-supervised learning with sklearn.semi_supervised, a module offering algorithms for training models on partially labeled data. Whether you’re facing limited labeled samples or seeking to leverage unlabeled data to improve model performance, this module provides innovative techniques for incorporating both labeled and unlabeled examples into the learning process. From self-training to co-training, sklearn.semi_supervised offers a range of algorithms suited to different problem domains and data distributions, empowering you to build more robust and accurate machine learning models with limited labeled data.

Support Vector Machines Estimators

Unleash the potential of Support Vector Machines (SVMs) with sklearn.svm, a module offering algorithms for classification, regression, and outlier detection. Whether you’re performing binary classification or multiclass classification, this module provides efficient and scalable solutions for separating data points into distinct classes while maximizing the margin of separation. With support for various kernel functions and optimization techniques, sklearn.svm offers flexible algorithms capable of handling nonlinear decision boundaries and complex data distributions, empowering you to build accurate and reliable predictive models across diverse machine learning tasks and domains.

Decision Tree Algorithms and Plotting Utilities

Traverse the branches of decision trees with sklearn.tree, a versatile module offering algorithms for building, visualizing, and interpreting decision tree models. Whether you’re performing classification or regression, decision trees offer intuitive and interpretable solutions for making predictions based on a series of simple rules. From pruning and feature selection to ensemble methods like random forests and gradient boosting, sklearn.tree provides a range of techniques for improving the accuracy and robustness of decision tree models across diverse machine learning tasks and domains.

Utility Functions for Input Validation and more

The sklearn.utils module provides a comprehensive set of utilities tailored to various aspects of machine learning workflows. It includes tools for input and parameter validation, ensuring the robustness and reliability of model training. Additionally, it offers utilities used in meta-estimators, facilitating the integration of multiple learning algorithms into composite models. The module also provides functionality for handling weights based on class labels, enabling the customization of model behavior to account for class imbalances. Moreover, it offers utilities to effectively handle multiclass targets in classifiers, ensuring accurate predictions across diverse classification tasks.

Furthermore, sklearn.utils includes tools for performing optimal mathematical operations, enhancing computational efficiency during model training and evaluation. It also provides functionality for working with sparse matrices and arrays, optimizing memory usage and computational performance for large-scale datasets. Additionally, utilities for graph manipulation, random sampling, array operations, metadata routing, object discovery, compatibility checking, and parallel computing are available, empowering users to streamline and accelerate various aspects of the machine learning pipeline.

Experimental Features and Functions

Sklearn.experimental will let you experiment with new features and experimental functionalities. With new algorithms and emerging techniques, this module offers a sneak peek into the future of machine learning, allowing you to stay ahead of the curve and push the boundaries of what’s possible in data science.

Exception Handling and Warnings

Use sklearn.exceptions for exception handling and warnings. There, you can handle errors gracefully and troubleshoot issues effectively. Whether you’re debugging code or optimizing performance, this module provides a comprehensive framework for managing exceptions and warnings, empowering you to write robust and reliable machine learning applications.