Microsoft open source model-based machine learning framework Infer.NET

On the 6th, Microsoft opened up its model-based machine learning framework, Infer.NET. Infer.NET is a framework for running Bayesian inference in a graphical model, and it can also be used for probabilistic programming. Infer.NET can be used to solve many different types of machine learning problems, including standard questions such as classification, recommendation, or clustering, and custom solutions for domain-specific issues. Infer.NET is now widely used in a variety of fields, including information retrieval, bioinformatics, epidemiology, vision, and many others.

The Infer.NET project was launched in 2004 by a team at the Microsoft Research Center in Cambridge, England, and was released for academic use in 2008. In Microsoft’s new world of AI, the technology has evolved into a machine learning engine and the game applications on Office and Azure and the Xbox.

Using a model-based approach to machine learning, developers provide models for the framework, and then the framework generates machine learning algorithms directly from the presented models. Many learning models require programmers to map their models to pre-existing learning algorithms. However, Infer.NET is the reverse process, which is the advantage of Infer.NET. Developers believe that as artificial intelligence software becomes more popular, interpreting system behaviour becomes more and more important, and users should be able to find out why the system behaves in some way given a model.

Infer.NET is cross-platform and supports .NET Framework 4.6.1, .NET Core 2.0, and Mono 5.0. Windows users can use it in Visual Studio 2017, and macOS and Linux people can integrate it with the command line option. In the code manager.