Specifically designed for large dimensions and a widely used special case of the Metropolis Algorithm.
- Initialize probabilities randomly or uniformly
- In each step, replace the value of one of the variables by a value drawn from the distribution of that variable conditioned on the values of the remaining variables
- Repeat until convergence
A technique for generating samples from complex multidimensional probability distributions by sampling from the joint distribution of all variables given their conditional distributions. For each variable, it repeatedly generates new samples by drawing from that variable's conditional distribution while keeping all other variables fixed.
- initialize x
- sample y based on
- sample x based on
- iterate and approximate distribution
- sample x, y