Microsoft has announced that they are making their Infer.NET framework open source for commercial applications.
Infer.NET is a machine learning framework that is used in a large number of Microsoft products like XBox, Azure, and Microsoft Office.
The Redmond-based company started the development of the framework in 2004 in their research lab in Cambridge and was aimed at developing scalable and interpretable machine learning solutions. By 2008, Microsoft made the tool available for academic use.
Infer.NET adopts a model-based approach to machine learning which allows developers to incorporate domain knowledge into the model. The model, then, can be used for building besp0oke machine learning algorithm with the help of the framework.
One of the most important features of Infer.NET is its interpretability.
“If you have designed the model yourself and the learning algorithm follows that model, then you can understand why the system behaves in a particular way or makes certain predictions.”
The team that developed Infer.NET has made the framework a part of ML.NET — the machine learning framework for .NET developers. Repositories have been set up under the .NET foundation, and this integration between the two frameworks will extend ML.NET for statistical modeling and online learning.
Those who are interested in Infer.NET can download the framework from here. The framework is supported on macOS, Windows, and Linux via .NET core. You can also go through Microsoft’s Tutorial and Examples pages before beginning with the framework. Developers can contribute their codes on Github’s Infer.NET repository.