Scikit-learn: Difference between revisions

Revision as of 20:46, 15 December 2022

Latest revision as of 19:06, 6 August 2025

(37 intermediate revisions by 29 users not shown)

Line 8:

| collapsible =

| author = [[David Cournapeau]]

| developer = [[Google Summer of Code]] project

| released = {{Start date and age|2007|06|df=yes}}

| latest release version = {{wikidata|property|reference|P348}}

Line 23:

| website = {{URL|https://scikit-learn.org/}}

}}

{{Portal|Free and open-source software}}

'''Scikit-learn''' (formerly '''scikits.learn''' and also known as '''sklearn''') is a [[free software]] [[machine learning]] [[Library (computing)|library]] for the [[Python (programming language)|Python]] [[programming language]].<ref name="jmlr">{{cite journal

'''scikit-learn''' (formerly '''scikits.learn''' and also known as '''sklearn''') is a [[free and open-source]] [[machine learning]] [[Library (computing)|library]] for the [[Python (programming language)|Python]] [[programming language]].<ref name="jmlr">{{cite journal

|author1=Fabian Pedregosa

|author2=Gaël Varoquaux

Line 39 ⟶ 40:

|author14=Matthieu Perrot

|author15=Édouard Duchesnay

|title=Scikitscikit-learn: Machine Learning in Python

|journal=Journal of Machine Learning Research

|year=2011

|volume=12

|pages=2825–2830

|arxiv=1201.0490

|bibcode=2011JMLR...12.2825P

|url=http://jmlr.org/papers/v12/pedregosa11a.html

}}</ref>

It features various [[statistical classification|classification]], [[regression analysis|regression]] and [[Cluster analysis|clustering]] [[Algorithm|algorithms]] including [[support vector machine|support-vector machine]]s, [[random forests]], [[gradient boosting]], [[k-means clustering|''k''-means]] and [[DBSCAN]], and is designed to interoperate with the [[Python (programming language)|Python]] numerical and scientific libraries [[NumPy]] and [[SciPy]]. Scikit-learn is a [[NumFOCUS]] fiscally sponsored project.<ref>{{cite web|title=NumFOCUS Sponsored Projects|url=https://numfocus.org/sponsored-projects|publisher=NumFOCUS|access-date=2021-10-25}}</ref>

==Overview==

The scikit-learn project started as scikits.learn, a [[Google Summer of Code]] project by French [[data scientist]] [[David Cournapeau]]. ItsThe name stems fromof the notionproject thatderives itfrom its role isas a "SciKit"scientific (SciPytoolkit Toolkit)for machine learning", aoriginally separately-developed and distributed as a third-party extension to [[SciPy]].<ref>{{cite web

|url=https://scikits.appspot.com/scikit-learn

|title=scikit-learn

|last1=Dreijer

|first1=Janto

}}</ref> The original [[codebase]] was later rewritten by other [[Programmer|developers]].{{Who|date=March 2025}} In 2010, contributors Fabian Pedregosa, Gaël Varoquaux, Alexandre Gramfort and Vincent Michel, from the [[French Institute for Research in Computer Science and Automation]] in [[Plateau de Saclay|Saclay]], [[France]], took leadership of the project and released the first public version of the library on February 1, 2010.<ref>{{cite web|url=https://scikit-learn.org/stable/about.html#history|title=About us — scikit-learn 0.20.1 documentation|website=scikit-learn.org}}</ref> In November 2012, scikit-learn as well as [[scikit-image]] were described as two of the "well-maintained and popular" {{As of|2012|11|alt=scikits libraries}}.<ref>{{cite book

}}</ref>

The original [[codebase]] was later rewritten by other developers. In 2010 Fabian Pedregosa, Gael Varoquaux, Alexandre Gramfort and Vincent Michel, all from the [[French Institute for Research in Computer Science and Automation]] in [[Plateau de Saclay|Saclay]], [[France]], took leadership of the project and made the first public release on February the 1st 2010.<ref>{{cite web|url=https://scikit-learn.org/stable/about.html#history|title=About us — scikit-learn 0.20.1 documentation|website=scikit-learn.org}}</ref> Of the various scikits, scikit-learn as well as [[scikit-image]] were described as "well-maintained and popular" {{As of|2012|11|alt=in November 2012}}.<ref>{{cite book

|author=Eli Bressert

|title=SciPy and NumPy: an overview for developers

Line 62 ⟶ 64:

|url=https://books.google.com/books?id=fLKTuJqQLVEC&pg=PA43

|page=43

|isbn=978-1-4493-6162-4

}}</ref> Scikit-learn is one of the most popular machine learning libraries on [[GitHub]].<ref>{{Cite web|url=https://github.blog/2019-01-24-the-state-of-the-octoverse-machine-learning/|title=The State of the Octoverse: machine learning|date=2019-01-24|website=The GitHub Blog|publisher=[[GitHub]]|language=en-US|access-date=2019-10-17}}</ref>

}}</ref> In 2019, it was noted that scikit-learn is one of the most popular machine learning libraries on [[GitHub]].<ref>{{Cite web|url=https://github.blog/2019-01-24-the-state-of-the-octoverse-machine-learning/|title=The State of the Octoverse: machine learning|date=2019-01-24|website=The GitHub Blog|publisher=[[GitHub]]|language=en-US|access-date=2019-10-17}}</ref>

== Features ==

* Large catalogue of well-established machine learning algorithms and data pre-processing methods (i.e. [[feature engineering]])

* Utility methods for common data-science tasks, such as splitting data into [[Training, validation, and test data sets|train and test sets]], [[Cross-validation (statistics)|cross-validation]] and [[grid search]]

* Consistent way of running machine learning models ({{code|estimator.fit()|python}} and {{code|estimator.predict()|python}}), which libraries can implement

* Declarative way of structuring a data science process (the {{Code|Pipeline|Python}}), including data pre-processing and model fitting

== Examples ==

Fitting a [[Random forest|random forest classifier]]:<syntaxhighlight lang="python3" line="1">

>>> from sklearn.ensemble import RandomForestClassifier

>>> classifier = RandomForestClassifier(random_state=0)

>>> X = [[ 1, 2, 3], # 2 samples, 3 features

... [11, 12, 13]]

>>> y = [0, 1] # classes of each sample

>>> classifier.fit(X, y)

RandomForestClassifier(random_state=0)

</syntaxhighlight>

==Implementation==

Scikitscikit-learn is largely written in Python, and uses [[NumPy]] extensively for high-performance linear algebra and array operations. Furthermore, some core algorithms are written in [[Cython]] to improve performance. Support vector machines are implemented by a Cython wrapper around [[LIBSVM]]; logistic regression and linear support vector machines by a similar wrapper around [[LIBLINEAR]]. In such cases, extending these methods with Python may not be possible.

scikit-learn integrates well with many other Python libraries, such as [[Matplotlib]] and [[plotly]] for plotting, [[NumPy]] for array vectorization, [[Pandas (software)|Pandas]] dataframes, [[SciPy]], and many more.

== History ==

scikit-learn was initially developed by David Cournapeau as a Google Summer of Code project in 2007. Later that year, Matthieu Brucher joined the project and started to use it as a part of his thesis work. In 2010, [[French Institute for Research in Computer Science and Automation|INRIA]], the [[French Institute for Research in Computer Science and Automation]], got involved and the first public release (v0.1 beta) was published in late January 2010.

== Applications ==

Scikit-learn is widely used across industries for a variety of machine learning tasks such as classification, regression, clustering, and model selection. The following are real-world applications of the library:

=== Finance and Insurance ===

* '''AXA''' uses scikit-learn to speed up the compensation process for car accidents and to detect insurance fraud.<ref name="sklearn-testimonials">{{Cite web |title=Testimonials |url=https://scikit-learn.org/stable/testimonials/testimonials.html |website=scikit-learn.org |access-date=2025-08-06}}</ref>

* '''Zopa''', a peer-to-peer lending platform, employs scikit-learn for credit risk modelling, fraud detection, marketing segmentation, and loan pricing.<ref name="sklearn-testimonials"/>

* '''BNP Paribas Cardif''' uses scikit-learn to improve the dispatching of incoming mail and manage internal model risk governance through pipelines that reduce operational and overfitting risks.<ref name="sklearn-testimonials"/>

* '''J.P. Morgan''' reports broad usage of scikit-learn across the bank for classification tasks and predictive analytics in financial decision-making.<ref name="sklearn-testimonials"/>

=== Retail and E-Commerce ===

* '''Booking.com''' uses scikit-learn for hotel and destination recommendation systems, fraudulent reservation detection, and workforce scheduling for customer support agents.<ref name="sklearn-testimonials"/>

* '''HowAboutWe''' uses it to predict user engagement and preferences on a dating platform.<ref name="sklearn-testimonials"/>

* '''Lovely''' leverages the library to understand user behaviour and detect fraudulent activity on its platform.<ref name="sklearn-testimonials"/>

* '''Data Publica''' uses it for customer segmentation based on the success of past partnerships.<ref name="sklearn-testimonials"/>

* '''Otto Group''' integrates scikit-learn throughout its data science stack, particularly in logistics optimization and product recommendations.<ref name="sklearn-testimonials"/>

=== Media, Marketing, and Social Platforms ===

* '''Spotify''' applies scikit-learn in its recommendation systems.<ref name="sklearn-testimonials"/>

* '''Betaworks''' uses the library for both recommendation systems (e.g., for Digg) and dynamic subspace clustering applied to weather forecasting data.<ref name="sklearn-testimonials"/>

* '''PeerIndex''' used scikit-learn for missing data imputation, tweet classification, and community clustering in social media analytics.<ref name="sklearn-testimonials"/>

* '''Bestofmedia Group''' employs it for spam detection and ad click prediction.<ref name="sklearn-testimonials"/>

* '''Machinalis''' utilizes scikit-learn for click-through rate prediction and relational information extraction for content classification and advertising optimization.<ref name="sklearn-testimonials"/>

* '''Change.org''' applies scikit-learn for targeted email outreach based on user behaviour.<ref name="sklearn-testimonials"/>

=== Technology ===

* '''AWeber''' uses scikit-learn to extract features from emails and build pipelines for managing large-scale email campaigns.<ref name="sklearn-testimonials"/>

* '''Solido''' applies it to semiconductor design tasks such as rare-event estimation and worst-case verification using statistical learning.<ref name="sklearn-testimonials"/>

* '''Evernote''', '''Dataiku''', and other tech companies employ scikit-learn in prototyping and production workflows due to its consistent API and integration with the Python ecosystem.<ref name="sklearn-testimonials"/>

=== Academia ===

* '''Télécom ParisTech''' integrates scikit-learn in hands-on coursework and assignments as part of its machine learning curriculum.<ref name="sklearn-testimonials"/>

Scikit-learn integrates well with many other Python libraries, such as [[Matplotlib]] and [[plotly]] for plotting, [[NumPy]] for array vectorization, [[Pandas (software)|Pandas]] dataframes, [[SciPy]], and many more.

== Version historyAwards ==

Scikit-learn was initially developed by David Cournapeau as a [[Google]] summer of code project in 2007. Later Matthieu Brucher joined the project and started to use it as a part of his thesis work. In 2010 [[French Institute for Research in Computer Science and Automation|INRIA]], the [[French Institute for Research in Computer Science and Automation]], got involved and the first public release (v0.1 beta) was published in late January 2010.

* August 2013. scikit-learn 0.14<ref name=":0" />

* July 2014. scikit-learn 0.15.0<ref name=":0" />

* March 2015. scikit-learn 0.16.0<ref name=":0" />

* November 2015. scikit-learn 0.17.0<ref name=":0">{{Cite web|url=https://scikit-learn.org/dev/whats_new.html|title=Release history — scikit-learn 0.19.dev0 documentation|website=scikit-learn.org|access-date=2017-02-27}}</ref>

* September 2016. scikit-learn 0.18.0

* July 2017. scikit-learn 0.19.0

* September 2018. scikit-learn 0.20.0<ref>{{cite web |title=Release History - 0.20.0 documentation |url=https://scikit-learn.org/stable/whats_new.html#version-0-20 |website=scikit-learn |access-date=6 November 2018}}</ref>

* May 2019. scikit-learn 0.21.0<ref>{{cite web |title=Release History - 0.21.0 documentation |url=https://scikit-learn.org/stable/whats_new.html#version-0-21-0 |website=scikit-learn |access-date=5 May 2019}}</ref>

* December 2019. scikit-learn 0.22.0<ref>{{cite web |title=Release History - 0.22.0 documentation |url=https://scikit-learn.org/dev/whats_new/v0.22.html#version-0-22-0 |website=scikit-learn |access-date=7 June 2020}}</ref>

*May 2020. scikit-learn 0.23.0<ref>{{cite web |title=Release History - 0.23.0 documentation |url=https://scikit-learn.org/dev/whats_new/v0.23.html#version-0-23-0 |website=scikit-learn |access-date=7 June 2020}}</ref>

* Jan 2021. scikit-learn 0.24<ref>{{Citation|title=scikit-learn: A set of python modules for machine learning and data mining|url=http://scikit-learn.org/|access-date=2021-02-08}}</ref>

* September 2021. scikit-learn 1.0<ref>{{Citation|title=scikit-learn: A set of python modules for machine learning and data mining|url=http://scikit-learn.org/|access-date=2021-09-24}}</ref>

* 2019 Inria-French Academy of Sciences-Dassault Systèmes Innovation Prize<ref>{{Cite web |title=The 2019 Inria-French Academy of Sciences-Dassault Systèmes Innovation Prize : scikit-learn , a success story for machine learning free software {{!}} Inria |url=https://www.inria.fr/en/2019-inria-french-academy-sciences-dassault-systemes-innovation-prize-scikit-learn-success-story |access-date=2025-03-19 |website=www.inria.fr}}</ref>

==Scikit-learn tools==

* 2022 Open Science Award for Open Source Research Software<ref>{{Cite web |last=Badolato |first=Anne-Marie |date=2022-02-07 |title=Open Science Awards for Open Source Research Software |url=https://www.ouvrirlascience.fr/open-science-free-software-award-ceremony/ |access-date=2025-03-19 |website=Ouvrir la Science |language=en}}</ref>

* [[mlpy]]

* [[SpaCy]]

* [[Natural Language Toolkit|NLTK]]

* [[Orange (software)|Orange]]

* [[PyTorch]]

* [[TensorFlow]]

* [[Infer.NET]]

* [[List of numerical analysis software]]

==References==

Line 102 ⟶ 147:

{{SciPy ecosystem}}

{{differentiable computing}}

[[Category:Data mining and machine learning software]]

Line 107 ⟶ 153:

[[Category:Python (programming language) scientific libraries]]

[[Category:Software using the BSD license]]

[[Category:2010 in artificial intelligence]]

[[Category:2010 software]]