Mandoline - Model Evaluation under Distribution Shift

Mandoline - Model Evaluation under Distribution Shift

Stanford MedAI via YouTube Direct link

Summary

14 of 20

14 of 20

Summary

Class Central Classrooms beta

YouTube playlists curated by Class Central.

Classroom Contents

Mandoline - Model Evaluation under Distribution Shift

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Intro
  2. 2 Outline
  3. 3 The ML model development process
  4. 4 Model Evaluation
  5. 5 Motivation
  6. 6 Common approach: importance weighting
  7. 7 Motivating example
  8. 8 Mandoline: Slice-based reweighting framework
  9. 9 The theory behind using slices
  10. 10 More formally...
  11. 11 Density Ratio Estimation
  12. 12 Experiments: tasks
  13. 13 Experiments: compare to reweighting on x
  14. 14 Summary
  15. 15 Taking a step back - how do we get slices? What are sli
  16. 16 Measuring model performance
  17. 17 Hidden Stratification: Approach
  18. 18 ML model development process, revisited
  19. 19 Another angle - how else can we evaluate?
  20. 20 "Closing the loop" - how do we update?

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.