| 贝叶斯网结构学习的研究现状及发展趋势 |
| |
| 引用本文： | 马壮,杨善林,胡小建.贝叶斯网结构学习的研究现状及发展趋势[J].合肥工业大学学报(自然科学版),2005,28(8):833-838. |
| |
| 作者姓名： | 马壮 杨善林 胡小建 |
| |
| 作者单位： | 合肥工业大学,管理学院,安徽,合肥,230009 |
| |
| 基金项目： | 国家自然科学基金资助项目(70471046)，教育部博士点基金资助项目(20040359004) |
| |
| 摘 要： | 目前,在结构已知情况下,贝叶斯网的参数学习算法及数据完备时的贝叶斯网的结构学习算法比较成熟,但是从不完全数据中学习贝叶斯网结构比较困难;文章简要介绍前者,重点分析了在不完备数据条件下结构学习的难点,对现有的学习算法进行了深入的研究和比较,对该领域的研究趋势进行了展望。
|
| 关 键 词： | 贝叶斯网 结构学习 不完备数据 算法 |
| 文章编号： | 1003-5060(2005)08-0833-06 |
| 修稿时间： | 2004年10月27 |
Research state and future trend of Bayesian network structure learning |
| |
| MA Zhuang,YANG Shan-lin,HU Xiao-jian.Research state and future trend of Bayesian network structure learning[J].Journal of Hefei University of Technology(Natural Science),2005,28(8):833-838. |
| |
| Authors: | MA Zhuang YANG Shan-lin HU Xiao-jian |
| |
| Abstract: | Bayesian networks(BN) that are natural compact representation of joint probability distribution have achieved remarkable results in the uncertainty community,and the researches in learning parameters of BN and learning structure of BN from complete data are successful,but learning structure of BN from incomplete data is a difficulty. In this paper,the former researches are described briefly,the difficulty of structure learning from incomplete data is analyzed emphatically,and a detailed discussion is made on the existing study algorithms. The research direction is also discussed. |
| |
| Keywords: | Bayesian network structure learning incomplete data algorithm |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
