Macroscopic Models of Clique Tree Growth for Bayesian Networks

In clique tree clustering, inference consists of propagation in a clique tree compiled from a Bayesian network. In this paper, we develop an analytical approach to characterizing clique tree growth as a function of increasing Bayesian network connectedness, specifically: (i) the expected number of moral edges in their moral graphs or (ii) the ratio of the number of non-root nodes to the number of root nodes. In experiments, we systematically increase the connectivity of bipartite Bayesian networks, and find that clique tree size growth is well-approximated by Gompertz growth curves. This research improves the understanding of the scaling behavior of clique tree clustering, provides a foundation for benchmarking and developing improved BN inference algorithms, and presents an aid for analytical trade-off studies of tree clustering using growth curves. **Reference:** O. J. Mengshoel, "Macroscopic Models of Clique Tree Growth for Bayesian Networks." In Proc. of the 22nd National Conference on Artificial Intelligence (AAAI-07). July 2007, Vancouver, Canada, pp. 1256-1262. **BibTex Reference:** @inproceedings{mengshoel07macroscopic, author = "Mengshoel, O. J.", title = "Macroscopic Models of Clique Tree Growth for {Bayesian} Networks", year = "2007", booktitle = {Proceedings of the Twenty-Second National Conference on Artificial Intelligence (AAAI-07)}, pages = "1256-1262", address = "Vancouver, British Columbia" }

                                            [
  "026:00"
]
                                        

                                            {
  "fn": "Ole Mengshoel",
  "@type": "vcard:Contact",
  "hasEmail": "mailto:ole.j.mengshoel@nasa.gov"
}
                                        

                                            [
  {
    "@type": "dcat:Distribution",
    "title": "GrowthV26aaai.pdf",
    "format": "PDF",
    "mediaType": "application/pdf",
    "description": "GrowthV26aaai.pdf",
    "downloadURL": "https://c3.nasa.gov/dashlink/static/media/publication/GrowthV26aaai.pdf"
  }
]
                                        

                                            [
  "ames",
  "dashlink",
  "nasa"
]
                                        

                                            [
  "026:029"
]
                                        

                                            {
  "name": "Dashlink",
  "@type": "org:Organization"
}