Conceptual
Login

Gibbs Sampling

Cycles through parameters, drawing each from its full conditional distribution given the current values of all others; every update is a Metropolis-Hastings step that is always accepted, but strong posterior correlations make the coordinate-wise moves mix slowly.

This Concept is waiting for its first lesson!

Cycles through parameters, drawing each from its full conditional distribution given the current values of all others; every update is a Metropolis-Hastings step that is always accepted, but strong posterior correlations make the coordinate-wise moves mix slowly.

Are you a teacher? Sign in to start contributing.

Sign In