Lessons and Outlook for ML Parameterization of Sub Grid Atmospheric Physics From the Vantage of Emulating Cloud Superparameterization - Mike Pritchard

Lessons and Outlook for ML Parameterization of Sub Grid Atmospheric Physics From the Vantage of Emulating Cloud Superparameterization - Mike Pritchard

Kavli Institute for Theoretical Physics via YouTube Direct link

Introduction

1 of 39

1 of 39

Introduction

Class Central Classrooms beta

YouTube playlists curated by Class Central.

Classroom Contents

Lessons and Outlook for ML Parameterization of Sub Grid Atmospheric Physics From the Vantage of Emulating Cloud Superparameterization - Mike Pritchard

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

  1. 1 Introduction
  2. 2 Motivation
  3. 3 Turbulence
  4. 4 Global modeling
  5. 5 The challenge
  6. 6 Multiskill modeling
  7. 7 Global storm resolving models
  8. 8 A silly first attempt
  9. 9 Aerosol cloud indirect effects
  10. 10 Regionalization
  11. 11 GPU Computing
  12. 12 Creative Complexity
  13. 13 Short Simulations
  14. 14 Course Graining
  15. 15 Super Crude Architecture
  16. 16 Lessons emerging
  17. 17 Feature engineering
  18. 18 Separate processes
  19. 19 Microphysical rates
  20. 20 Example
  21. 21 Constraints
  22. 22 Tradeoffs
  23. 23 Generalization
  24. 24 Strategy
  25. 25 Preprint
  26. 26 Results
  27. 27 Physical Credibility
  28. 28 Hyperparameter Tuning
  29. 29 Missing Information
  30. 30 Neural Network Tuning
  31. 31 Summary
  32. 32 Cognitive dissonance
  33. 33 Excitement
  34. 34 Thank you
  35. 35 Maria
  36. 36 Reporting failures
  37. 37 Retraining neural networks
  38. 38 Sampling
  39. 39 Failures

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.