AI engineering glossary · Classical ML algorithms
What is expectation-maximisation?
Expectation-maximisation (also called EM algorithm): An iterative method for fitting models with hidden variables. The E-step estimates the hidden values given current parameters, and the M-step updates parameters to best fit those estimates. Each cycle never lowers the likelihood, and it is used for mixture models.
Expectation-maximisation: an example
Fitting a two-component mixture: estimate each point's probability of belonging to component A, then recompute A's mean and spread using those probabilities.
Related terms
More Classical ML algorithms terms
- Explained variance ratio
- Extremely randomised trees
- Feature importance
- Gaussian mixture model
- Gaussian naive Bayes
- Gaussian process
- Gini impurity
- Gradient boosting
Where you learn it
Expectation-maximisation is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.