An Alternative Proof That Exact Inference Problem in Bayesian Belief Networks Is NP-Hard

摘要

Exact inference problem in belief networks has been well studied in the literature and has various application areas. It is known that this problem and its approximation version are NP-hard. In this study, an alternative polynomial time transformation is provided from the well-known vertex cover problem. This new transformation may lead to new insights and polynomially solvable classes of the exact inference problem in Bayesian belief networks.