ABSTRACT
One simple approximation scheme is to truncate evaluation of the integral∫ y∈X V(y)f (y |x, a)dµ. This would be especially advantageous if the support of density f (y |x, a) was large, hence computationally burdensome, while concentrating most of the probability mass onto a relatively small subset E⊂X . It would then be expected that a good approximation could be obtained by restricting evaluation of∫ y∈X V(y)f (y |x, a)dµ to E, while significantly reducing the computation time. The material in this chapter is based on Almudevar and de Arruda (2012) and Arruda et al. (2013).
