Owned By Statistics - How Kubeflow & MLOps Can Help Secure Your ML Workloads

Owned By Statistics - How Kubeflow & MLOps Can Help Secure Your ML Workloads

CNCF [Cloud Native Computing Foundation] via YouTube Direct link

Introduction

1 of 43

1 of 43

Introduction

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Owned By Statistics - How Kubeflow & MLOps Can Help Secure Your ML Workloads

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

  1. 1 Introduction
  2. 2 Machine Learning at Microsoft
  3. 3 ML in every product at Microsoft
  4. 4 ML in the average enterprise
  5. 5 Data scientist
  6. 6 Building a model
  7. 7 Rolling it out
  8. 8 Security
  9. 9 Three types of attacks
  10. 10 Advanced models
  11. 11 Snow detection
  12. 12 Stop sign detection
  13. 13 Face recognition
  14. 14 Defend against adversaries
  15. 15 Build an MLOps pipeline
  16. 16 Modular components
  17. 17 Pipeline example
  18. 18 Another attack vector
  19. 19 Malicious users
  20. 20 Two types of attacks
  21. 21 Distillation attack
  22. 22 Accuracy
  23. 23 GoogleBERT
  24. 24 Continuous Improvement
  25. 25 Build Efficient Pipelines
  26. 26 Take Your Models
  27. 27 Hidden Data
  28. 28 Recommendations
  29. 29 Network Graph
  30. 30 Map Leakage
  31. 31 Example
  32. 32 How to prevent this
  33. 33 Injections
  34. 34 Leaks
  35. 35 Summary
  36. 36 The Reality
  37. 37 You will be attacked
  38. 38 Conclusion
  39. 39 Questions
  40. 40 Reprocessing ML Pipeline Predictions
  41. 41 MLOps vs Continuous Machine Learning
  42. 42 Regulation of ML
  43. 43 Mitigating Leaky Data

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.