Updating Probabilities in Multiply-Connected Belief Networks
This work addresses a computational bottleneck for researchers and practitioners using belief networks, but it is incremental as it builds on existing methods without introducing a new paradigm.
The paper tackles the problem of probability updates in multiply-connected belief networks by proposing a heuristic algorithm to find a loop-cutset that meets specific conditions, enabling the application of Pearl's conditioning method to such networks.
This paper focuses on probability updates in multiply-connected belief networks. Pearl has designed the method of conditioning, which enables us to apply his algorithm for belief updates in singly-connected networks to multiply-connected belief networks by selecting a loop-cutset for the network and instantiating these loop-cutset nodes. We discuss conditions that need to be satisfied by the selected nodes. We present a heuristic algorithm for finding a loop-cutset that satisfies these conditions.