Machine learning for C# developers: Introducing ML
summary
Claims that ML.NET lets C# developers train and operationalize real ML without Python, using a dataview pipeline, 80/20 train/test split, one‑hot encoding and featureize‑text transforms. Demos include a Zillow regression (≈440 rows) with fast forest regression and R² ≈ 0.74 (cross‑validation lower), sentiment analysis on 1,000 Yelp reviews yielding near‑0/1 probabilities, Enron spam classification, and transfer learning of Google’s Inception with ~20 images to build a working hot‑dog WPF app.
AI-written summary · about
chapters (9)
- 0:00Introduction and ML revolution
- 10:05Understanding ML.NET and DataView
- 22:05Building regression models
- 33:29Training and evaluating models
- 43:31Classification models
- 48:50Sentiment analysis demo
- 52:40Spam detection demo
- 55:29Image classification with transfer learning
- 1:01:10Conclusion and resources
related talks
59mMachine Learning in .NET (ML.NET)
52mML.NET - Machine Learning made for .NET
57mMachine Learning for .NET Developers with ML.NET
57mMachine Learning Made Easy – An Introduction to ML.NET
01h 04mIntroduction to ML.NET
59mMachine Learning simplified for Developers with ML.NET
58mExploring Machine Learning for .Net Developers
57mMachine Learning for .Net Developers