AI Rookies

Gibbs Sampling

Fact

An MCMC method for sampling one variable at a time, using the others.

In Plain Words

Gibbs Sampling is like a potluck group chat. It keeps the menu, then changes one sad salad after checking the other dishes.

It updates one unknown at a time in Bayesian models and graph models. Exact probability is too hard there.

Related Concepts

MCMC
Gibbs Sampling is an MCMC method using conditional samples to approach a target distribution.

PGM
In a PGM, Gibbs Sampling estimates hard posterior probabilities.

Bayesian Network
In a Bayesian Network, Gibbs Sampling samples each variable from its conditional probability.

Boltzmann Machine
Boltzmann Machine training often uses Gibbs Sampling to draw energy-based samples.