[an error occurred while processing the directive]
RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

AutoGluon (library for machine learning)

Product
The name of the base system (platform): Amazon Web Services (AWS)
Developers: Amazon
Date of the premiere of the system: January, 2020
Branches: Information technologies
Technology: Development tools of applications

2020: Announcement

In January, 2020 the Amazon company released library open source which will help developers to create quickly applications using machine learning. The project under the name AutoGluon is created in division of Amazon Web Services (AWS).

Thanks to AutoGluon it is possible to create the models of artificial intelligence using images, the text and tabular data sets by writing of a small amount of code lines.

Amazon released open library for creation of AI models several code lines

Creation of such applications — a task difficult as developers should know how to configure "hyper parameters" at creation of AI model. They need also to deal with such problems as search of neural architecture which allows them to find the best architectural option for the models of machine learning. AutoGluon automates accomplishment of such tasks.

File:Aquote1.png
We developed AutoGluon really to democratize machine learning and to give an opportunity of deep learning to all developers" — the application AWS application programs specialist Jonas Mueller says. - Because of the opaque nature of deep learning many of solutions of experts are based on an intuition, but not on strict scientific understanding of how the individual choice affects desirable results" — Mueller notes. — AutoGluon solves this problem as all options are automatically configured within ranges by default which, as we know, well work for a specific objective and model.
File:Aquote2.png

AutoGluon is constructed on the basis of Gluon ML library for creation of models of a set of the ready optimized elements. The project is compatible in Python 3.6 and 3.7 and so far supports only Linux. The release of the version for macOS and Windows is planned, but terms are not specified.[1]

Notes