Authors: Yifeng Zeng, Yanping Xiang, Saulius Pacekajus
Addresses: Department of Computer Science, Aalborg University, Aalborg, Denmark. ' Department of Computer Science, University of Electronic Science and Technology of China, Chengdu, PR China. ' Department of Computer Science, Aalborg University, Aalborg, Denmark
Abstract: Refinement of Bayesian network (BN) structures using new data becomes more and more relevant. Some work has been done there; however, one problem has not been considered yet – what to do when new data have fewer or more attributes than the existing model. In both cases, data contain important knowledge and every effort must be made in order to extract it. In this paper, we propose a general merging algorithm to deal with situations when new data have different set of attributes. The merging algorithm updates sufficient statistics when new data are received. It expands the flexibility of BN structure refinement methods. The new algorithm is evaluated in extensive experiments and its applications are discussed at length.
Keywords: Bayesian networks; structure learning; sufficient statistics; new data; merging algorithm; attributes.
International Journal of Granular Computing, Rough Sets and Intelligent Systems, 2009 Vol.1 No.2, pp.203 - 220
Available online: 27 Aug 2009 *Full-text access for editors Access for subscribers Purchase this article Comment on this article