计算机科学 ›› 2015, Vol. 42 ›› Issue (4): 253-257.doi: 10.11896/j.issn.1002-137X.2015.04.052
肖 蒙,张友鹏
XIAO Meng and ZHANG You-peng
摘要： 针对贝叶斯网络中多父节点条件概率分布参数学习问题,提出了一种适用于多态节点、模型不精确、样本信息不充分情形的参数学习方法。该方法利用因果机制独立假设,分解条件概率分布,使条件概率表的规模表现为父节点个数和状态数的线性形式；利用Leaky Noisy-MAX模型量化了多态系统模型未含因素对参数学习的影响；从小样本数据集中获取模型参数并合成条件概率表。结果表明,该方法能提高参数学习效率与精度。
| [1] 张连文,郭海鹏.贝叶斯网引论[M].北京:科学出版社,2006 [2] Pearl J.Probabilistic Reasoning in Intelligent Systems:Net-works of Plausible Inference[M].San Francisco,CA:Morgan Kaufmann Publishers Inc.,1988:383-408 [3] 王华伟,周经伦,何祖玉,等.基于贝叶斯网络的复杂系统故障诊断[J].计算机集成制造系统,2004,0(2):230-234 [4] 康长青,方磊,华丽,等.基于贝叶斯Noisy Or Gate网络的多传感器目标分类识别[J].计算机测量与控制,2011,9(6):1387-1389 [5] 柴慧敏,王宝树.用于态势估计的一种构造贝叶斯网络参数的方法[J].计算机科学,2006,3(9):140-142 [6] Heckerman D.Causal Independence for Knowledge Acquisition and Inference[C]∥Proceedings of the 9th Conference on Uncertainty in Artificial Intelligence (UAI’93).San Mateo:Morgan Kaufmann Publishers Inc.,1993:122-127 [7] Zhang N L,Poole D.Exploiting Causal Independence in Bayesian Network Inference[J].Journal of Artificial Intelligence Research,1996,5:301-328 [8] Díez F J,Galán S F.Efficient Computation for the Noisy-Max[J].International Journal of Intelligent Systems,2004,8(2):165-177 [9] Díez F J,Druzdzel M J.Canonical Probabilistic Models forKnowledge Engineering[R].Technical Report CISIAD-06-01.UNED Madrid,2007 [10] Spirtes P,Glymour C N,Scheines R.Causation,Prediction,and Search[M].Cambridge:The MIT Press,2000 [11] Zagorecki A,Druzdzel M J.Knowledge Engineering for Bayesian Networks:How Common Are Noisy-MAX Distributions in Practice?[J].IEEE Transactions on Systems,Man,and Cybernetics:Systems,2013,3(1):186-195 [12] Shi D.Extending Noisy-Max Gates to Bidirectional Models[J].Journal of Information & Computational Science,2013,0(13):4085-4096 [13] Zagorecki A,Voortman M,Druzdzel M J.Decomposing LocalProbability Distributions in Bayesian Networks for Improved Inference and Parameter Learning[C]∥Proceedings of the 19th International Florida Artificial Intelligence Research Society Conference(FLAIRS-2006).Menlo Park,CA:AAAI Press,2006:860-865 |
| No related articles found! |
|
||
首页