Expectation-Maximization
Jump to navigation
Jump to search
Expectation-maximization (EM) is an iterative algorithm for finding estimates of parameters in probabilistic models, where the model contains unobserved latent variables, which is often the case in collaborative filtering.
