accord-net/framework
Accord.NET is a .NET machine learning framework providing algorithms for classification, regression, statistics, computer vision, and neural networks.

The framework offers implementations of support vector machines, neural networks, statistical analysis, and image processing for the .NET ecosystem. It includes components for machine learning model training, evaluation, and deployment within C# and other .NET languages. The library was designed to make ML techniques accessible to .NET developers and was used in various scientific computing applications over its 14-year development span.