| 具有丢失数据的贝叶斯网络结构学习算法 |
| |
| 引用本文： | 孙岩,唐一源.具有丢失数据的贝叶斯网络结构学习算法[J].计算机工程与设计,2008,29(1):142-143,250. |
| |
| 作者姓名： | 孙岩 唐一源 |
| |
| 作者单位： | 1. 辽宁师范大学计算机与信息技术学院,辽宁,大连,116029;大连理工大学神经信息学研究所,辽宁,大连,116023
2. 大连理工大学神经信息学研究所,辽宁,大连,116023 |
| |
| 摘 要： | 学习具有丢失数据的贝叶斯网络结构主要采用结合 EM 算法的打分一搜索方法,其效率和可靠性比较低.针对此问题建立一个新的具有丢失数据的贝叶斯网络结构学习算法.该方法首先用 Kullback-Leibler(KL)散度来表示同一结点的各个案例之间的相似程度,然后根据 Gibbs 取样来得出丢失数据的取值.最后,用启发式搜索完成贝叶斯网络结构的学习.该方法能够有效避免标准 Gibbs 取样的指数复杂性问题和现有学习方法存在的主要问题.
|
| 关 键 词： | KL散度 Gibbs取样 结构学习 丢失数据 贝叶斯网络 |
| 文章编号： | 1000-7024（2008）01-0142-02 |
| 收稿时间： | 2007-01-22 |
| 修稿时间： | 2007年1月22日 |
Learning Bayesian network structure with missing data |
| |
| SUN Yan,TANG Yi-yuan.Learning Bayesian network structure with missing data[J].Computer Engineering and Design,2008,29(1):142-143,250. |
| |
| Authors: | SUN Yan TANG Yi-yuan |
| |
| Abstract: | At present, the algorithm of learning Bayesian structure with missing data is mainly based on the search and scoring method combined with EM algorithm. The algorithm had low efficiency. A new algorithm of learning Bayesian network structure with missing data is presented. KL divergence is used to express the similarity between the cases. Then the value of the missing data is draw according to the Gibbs sampling. Finally, heuristic search is used to complete the learning of Bayesian network structure, This method can avoid the exponential complexity of standard Gibbs sampling and the main problems in the existing algorithm. |
| |
| Keywords: | Kullback-Leibler divergence Gibbs sampling structure learning missing data Bayesian network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
