Computing probability intervals with simulated annealing and probability trees
Andrés Cano, Juan Manuel Fernández-Luna, Serafı́n Moral · Journal of Applied Non-Classical Logics · 2002
This paper presents a method to compute a posteriori probability intervals when the initial conditional information is also given with probability intervals. The right way to make an exact computation is with the associated convex set of probabilities. Probability trees are used to represent these initial conditional convex sets because they greatly save the space required. This paper proposes a simulated annealing algorithm, which uses probability trees to represent the convex sets in order to compute the a posteriori intervals.