Best Practices in Machine Learning Prediction

 In this webinar, Vladyslav Khizhanov, discusses how to build optimal machine learning-based prediction models with our Auger.AI Automated Machine Learning Service.

The following topics are covered in the discussion:

  • Sizing your instances based on data size

  • Choosing algorithms (i.e. restricting search) based on data dimensionality (rows and columns) and data distribution (edited)

  • Restricting search based on explainability

  • Restricting search based on prediction runtime deployment size and speed

  • Choosing the appropriate accuracy metric to optimize for