Browse Teach Curate

Expectation-Maximization Algorithm in Statistical Learning

Fitting models with latent or missing variables by alternating an expectation step that computes the conditional distribution of the hidden variables with a maximisation step that reoptimises the parameters, and why each cycle cannot decrease the observed-data likelihood.

jeremy Video

Lecture 13 - Expectation-Maximization Algorithms | Stanford CS229: Machine Learning (Autumn 2018)

The Expectation-Maximization (EM) algorithm is a general iterative method for finding maximum likelihood estimates in statistical models with latent unobserved variables within the domain of unsuperv…