What do you want to see next in ML.NET?

ML.NET is an open source and cross-platform machine learning framework made for .NET developers.

Using ML.NET, you can stay in .NET to easily build and consume custom machine learning models for scenarios like sentiment analysis, price prediction, sales forecasting, recommendation, image classification, and more.

Over the past six months, the team has been working hard on fixing bugs, improving documentation, and adding more features and capabilities based on user feedback. This includes:

• Enhancements for .NET Core 3.0
• Azure training for image classification in Model Builder
• Expanded support for ONNX export
• Database loader for model training directly against relational databases
• Simplified Image Classification API for training image classification models
• Support for ML.NET in Jupyter Notebooks

Now we’d like to see how you’re using ML.NET and what features we can add and/or improve to make the framework and tooling even better.

Through the survey below, we would love to get feedback on how we can improve ML.NET. We will use your feedback to drive the direction of ML.NET and update our roadmap.

Thanks!

ML.NET team