sequential monte carlo (smc) methods

A class of methods used to sample from probability distributions based on constructing a sequence of intermediate distributions. Used in Bayesian neural networks to approximate posterior distributions.