Latent Variable Formulation of Mixture Models
Each observation x is paired with an unobserved one-hot indicator z choosing its generating component; marginalizing z out of the joint p(x, z) = p(z) p(x | z) recovers the mixture density and turns fitting into an incomplete-data inference problem.
This Concept is waiting for its first lesson!
Each observation x is paired with an unobserved one-hot indicator z choosing its generating component; marginalizing z out of the joint p(x, z) = p(z) p(x | z) recovers the mixture density and turns fitting into an incomplete-data inference problem.
Are you a teacher? Sign in to start contributing.
Sign In