Title :
Mapping and parallel implementation of Bayesian belief networks
Author :
Saxena, Nina ; Sarkar, Sudeep ; Ranganathan, N.
Author_Institution :
Dept. of Comput. Sci. & Eng., Univ. of South Florida, Tampa, FL, USA
Abstract :
Presents an efficient technique for mapping arbitrarily large Bayesian belief networks on hypercubes with deadlock-free implementation. We show that the speedup does not vary with the number of nodes in the Bayesian network and is limited by the height of the Peot-Shachter tree which is obtained by hanging the Bayesian polytree by a pivot node. We also found that the overhead in implementing Bayesian networks on parallel machines like hypercubes can be large because of the communication intensive nature of the network
Keywords :
hypercube networks; inference mechanisms; message passing; parallel algorithms; Bayesian belief networks; Peot-Shachter tree; communication intensive; deadlock-free; hypercubes; overhead; parallel machines; Artificial intelligence; Bayesian methods; Computer networks; Computer vision; Distributed computing; Hypercubes; Microelectronics; Random variables; System recovery; Visualization;
Conference_Titel :
Parallel and Distributed Processing, 1996., Eighth IEEE Symposium on
Conference_Location :
New Orleans, LA
Print_ISBN :
0-8186-7683-3
DOI :
10.1109/SPDP.1996.570391
