Learning Belief Networks in Domains with Recursively Embedded Pseudo Independent Submodels
This work addresses a specific issue in probabilistic domain modeling for researchers in machine learning and AI, representing an incremental improvement over prior methods.
The paper tackled the problem of learning belief networks in domains with recursively embedded pseudo independent submodels, which previous algorithms could miss, and proposed an improved algorithm that ensures learning of all such submodels up to a size bound with only a slight increase in complexity, as demonstrated experimentally.
A pseudo independent (PI) model is a probabilistic domain model (PDM) where proper subsets of a set of collectively dependent variables display marginal independence. PI models cannot be learned correctly by many algorithms that rely on a single link search. Earlier work on learning PI models has suggested a straightforward multi-link search algorithm. However, when a domain contains recursively embedded PI submodels, it may escape the detection of such an algorithm. In this paper, we propose an improved algorithm that ensures the learning of all embedded PI submodels whose sizes are upper bounded by a predetermined parameter. We show that this improved learning capability only increases the complexity slightly beyond that of the previous algorithm. The performance of the new algorithm is demonstrated through experiment.