Machine Learning — Medium
Key points
- EM algorithm alternates between E-step and M-step
- Specifically designed for models with hidden variables
- Goal is to maximize likelihood estimates
- Iterative process to converge to optimal parameters
Ready to go further?
Related questions
