Algorithms for Learning Decomposable Models and Chordal Graphs
This work addresses a specific problem in graphical models and graph theory, offering incremental algorithmic improvements for researchers in these domains.
The paper tackled the problem of learning decomposable models and chordal graphs by developing an exact algorithm to recover chordal graphical representations from decomposable models and proposing an algorithm for learning chordal approximations of general undirected graphs, but it did not provide concrete numerical results.
Decomposable dependency models and their graphical counterparts, i.e., chordal graphs, possess a number of interesting and useful properties. On the basis of two characterizations of decomposable models in terms of independence relationships, we develop an exact algorithm for recovering the chordal graphical representation of any given decomposable model. We also propose an algorithm for learning chordal approximations of dependency models isomorphic to general undirected graphs.