Outline

Content

• Machine Learning for Data Streams

  • Stream vs. batch learning
  • Challenges of stream learning
  • Non-stationary (evolving) streams – Concept drift
  • Multi-label learning

Materials: will be available after the tutorial.

• River

  • Overview
  • Running experiments
  • Extending/implementing estimators
  • Concept Drift

Materials: GitHub repository.

Additional resources

• River webpage: https://riverml.xyz

  • Documentation
  • Examples

• GitHub repository: https://github.com/online-ml/river/

  • Follow development
  • Interact with the development team and the project's community (Discussions)
  • Report issues and bugs
• River: machine learning for streaming data in Python (arXiv)