Clustering a Mixture of Gaussians with Unknown Covariance - Lecture

Clustering a Mixture of Gaussians with Unknown Covariance - Lecture

USC Probability and Statistics Seminar via YouTube Direct link

The problem today

2 of 22

2 of 22

The problem today

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Clustering a Mixture of Gaussians with Unknown Covariance - Lecture

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

  1. 1 Intro
  2. 2 The problem today
  3. 3 Challenges
  4. 4 How to measure separation?
  5. 5 Statistical metrics
  6. 6 Questions
  7. 7 Previous work: unknown covariance
  8. 8 Numerical illustration: FashionMNIST
  9. 9 Insight: Invariance
  10. 10 Canonical form
  11. 11 Maximum likelihood estimator
  12. 12 Optimality of Max-Cut
  13. 13 Two stage algorithm
  14. 14 Projected power iteration
  15. 15 Spectral algorithm
  16. 16 Global convergence guarantee
  17. 17 A statistical-computational gap?
  18. 18 A hard testing problem
  19. 19 Spectral methods lower bound
  20. 20 A reduction from testing
  21. 21 Max-Cut Semidefinite relaxation
  22. 22 Summary

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.