Conference Paper

Conditioning Graphs: Practical Structures for Inference in Bayesian Networks.

DOI: 10.1007/11589990_8 Conference: AI 2005: Advances in Artificial Intelligence, 18th Australian Joint Conference on Artificial Intelligence, Sydney, Australia, December 5-9, 2005, Proceedings
Source: DBLP

ABSTRACT Programmers employing inference in Bayesian networks typ- ically rely on the inclusion of the model as well as an inference engine into their application. Sophisticated inference engines require non-trivial amounts of space and are also difficult to implement. This limits their use in some applications that would otherwise benefit from probabilistic inference. This paper presents a system that minimizes the space require- ment of the model. The inference engine is sufficiently simpleas to avoid space-limitation and be easily implemented in almost any environment. We show a fast, compact indexing structure that is linear in the size of the network. The additional space required to compute over the model is linear in the number of variables in the network.

  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents adaptive hybrid protocols based on the declarative network and mainly discusses the principle and realization of the Bayesian-estimation based adaptive hybrid protocol in the declarative network, which is well adapted to the Mobile Ad hoc NETwork (MANET). The adaptive hybrid protocol is designed for ad hoc networks which have characteristics like self-organizing, no trusted party, flexibility, etc. The nodes that run the hybrid protocol can automatically select one routing protocol that is suitable for different network environment. The Bayesian-estimation based adaptive strategy, that improves the adaptability and stability of the protocol, succeeds in the Rapidnet, a declarative network engine. The result in the Rapidnet proves that the hybrid protocol and the adaptive strategy are feasible. The experiment on the ns-3 simulator, an emerging discrete-event network simulator, validates that this protocol performs well and reduces communication overheads.
    Journal of Electronics (China) 08/2013; 30(4). DOI:10.1007/s11767-013-3037-y
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A conditioning graph (CG) is a graphical structure that at- tempt to minimize the implementation overhead of computing probabil- ities in belief networks. A conditioning graph recursively factorizes the network, but restricting each decomposition to a single node allows us to store the structure with minimal overhead, and compute with a simple algorithm. This paper extends conditioning graphs with optimizations that effectively reduce the height of the CG, thus reducing time com- plexity exponentially, while increasing the storage requirements by only a constant factor. We conclude that CGs are frequently as efficient as any other exact inference method, with the advantage of being vastly superior to VE and JT in terms of space complexity, and far simpler to implement.
    Advances in Artificial Intelligence, 19th Conference of the Canadian Society for Computational Studies of Intelligence, Canadian AI 2006, Québec City, Québec, Canada, June 7-9, 2006, Proceedings; 01/2006
  • [Show abstract] [Hide abstract]
    ABSTRACT: We consider efficient indexing methods for conditioning graphs, which are a form of recursive decomposition for Bayesian networks. We compare two well-known methods for indexing, a top-down method and a bottom-up method, and discuss the redundancy that each of these suffer from. We present a new method for indexing that combines the advantages of each model in order to reduce this redundancy. We also introduce the concept of an update manager, which is a node in the conditioning graph that controls when other nodes update their current index. Empirical evaluations over a suite of standard test networks show a considerable reduction both in the amount of indexing computation that takes place, and the overall runtime required by the query algorithm.
    International Journal of Approximate Reasoning 10/2012; 53(7):969–987. DOI:10.1016/j.ijar.2012.03.001 · 1.73 Impact Factor

Preview (2 Sources)

Available from

Michael C. Horsch