Stochastic Sampling Algorithms: Self Importance Sampling
From DSL
The self-importance sampling algorithm is described in (Shachter & Peot 1990). It controls the generation of samples to account for bias due to disproportional sampling of cases that represent the most probable hypotheses.
