You will no doubt be familiar with the Python language, but did you know that it has a library dedicated to artificial intelligence? While doing some research, I found that this fascinating subject was not talked about enough. In this article, discover Scikit-learn, the Python library dedicated to machine learning.
Open-source machine learning
Are you a developer or a researcher looking to make use of various Python capabilities such as image recognition? That is now possible: if you know how the Python language works and have a solid grasp of English, you can use the Scikit-learn library free of charge.
Several example applications are explained on the library’s official website. The most surprising is still facial recognition: using a database filled with photos, it is possible to find the one that most closely matches a composite sketch.
But there are other applications too, such as recognising handwritten digits. By writing digits on a sheet of paper, your machine learns to digitise and recognise your handwriting.
A French success story!
Scikit-learn was created in 2007 by David Cournapeau, a student at the renowned engineering school Télécom ParisTech. The project is now supported by the French National Institute for Research in Digital Science and Technology (Inria). It is currently maintained by a new student from the school together with two engineers.
A library adopted by the most influential companies
Among the users of Scikit-learn are major multinationals such as:
- Spotify, which uses machine learning for its music recommendation engine
- Inria, which applies the library to several research programmes such as neuroimaging, computer vision and medical image analysis
- Evernote, which uses Scikit-learn for several processes such as data mining
- Booking.com, which uses the library to recommend hotels and destinations to its customers
