Extension for Scikit-learn* is happy to introduce 2026.1.0 release!

🔨 Library Engineering

• Added Sklearn 1.9 support for Extension for Scikit-learn* packages

🚨 What's New

• Introduced new Extension for Scikit-learn* functionality:
  • Introduced the “everything follows X” design pattern across multiple algorithms (linear models, KNN, decision forests, and basic statistics)
  • Implemented namespace and device checks for logistic regression
  • Added number of iterations to SVM models, improving training observability

🪲 Bug Fixes

• Fixed KMeans prediction on sparse data after dense training
• Restored proper data movement behavior in train_test_split
• Fixed BufferError when using DLPack capsules
• Fixed creation and handling of oneDAL objects after fallback paths in linear models and logistic regression
• Improved error handling and expanded error types for updated scikit-learn logic
• Fixed incorrect n_jobs restoration and warnings in newer scikit-learn versions

Acknowledgements

Thanks to everyone who helped us make 2026.1.0 release possible!

@Alexandr-Solovev, @Vika-F, @david-cortes-intel, @icfaust, @napetrov, @maria-Petrova, @homksei, @ahuber21, @ethanglaser, @razdoburdin, @avolkov-intel, @yuejiaointel, @DDJHB, @kjackiew, @richardnorth3

Full Changelog: 2026.0.0...2026.1.0