《Learning Bayesian network structure with immune algorithm》.pdfVIP

Journal of Systems Engineering and Electronics Vol. 26, No. 2, April 2015, pp.282–291 Learning Bayesian network structure with immune algorithm * Zhiqiang Cai, Shubin Si , Shudong Sun, and Hongyan Dui Ministry of Education Key Laboratory of Contemporary Design and Integrated Manufacturing Technology, School of Mechatronics, Northwestern Polytechnical University, Xi’an 710072, China Abstract: Finding out reasonable structures from bulky data is BN, and computing the corresponding parameters of con- one of the difficulties in modeling of Bayesian network (BN), which ditional probability distributions in BN [7]. The structure is also necessary in promoting the application of BN. This pa- of the BN reflects the underlying probabilistic dependence per proposes an immune algorithm based method (BN-IA) for the relations among the nodes and a set of assertions about learning of the BN structure with the idea of vaccination. Further- more, the methods on how to extract the effective vaccines from conditional independencies. local optimal structure and root nodes are also described in details. The main problem of learning the BN structure is to find Finally, the simulation studies are implemented with the helicopter the most reasonable network structure to extract all po- convertor BN model and the car start BN model. The comparison tential relationships from the dataset comprehensively. Be- results show that the proposed vaccines and the BN-IA can learn cause building the network structure of practical BN from the BN structure effectively and efficiently. data is an NP-hard problem