nested-sampling

Sampling bounded shells of the hypothesis space, in order to efficiently compute the Bayesian [[evidence]].

The best way I've heard it put:

MCMC is solving a hard problem once; nested sampling is solving an easier problem many times.