首页> 外文会议>Machine learning >Learning Bayesian Belief Networks Based on the Minimum Description Length Principle: An Efficient Algorithm Using the B B Technique
【24h】

Learning Bayesian Belief Networks Based on the Minimum Description Length Principle: An Efficient Algorithm Using the B B Technique

机译:基于最小描述长度原理的贝叶斯信念网络学习:使用B&B技术的高效算法

获取原文
获取原文并翻译 | 示例

摘要

In this paper, the problem of learning a Bayesian belief network (BBN) from given examples based on the minimum description length (MDL) principle is addressed. Given examples, the MDL-based procedure computes for each network the total of description length of the network and that of the examples given the network, and finds some network with the minimum value. We provide a search algorithm that reduces the computation time of the MDL-based procedure and at the same time is sure to find the network with the MDL. The proposed algorithm, which applies the branch and bound (B & B) technique, has lower computational complexity compared with the exhaustive search to the problem. Some empirical experiments by the Alarm database show that the proposed algorithm is fairly efficient for various problems of a moderate size. While the B & B technique has not been applied thus far to MDL-based/Bayesian procedures, the result is considered to suggest an advantage of the MDL-based procedure over the Bayesian procedure proposed by Cooper and Herskovits.
机译:在本文中,解决了基于最小描述长度(MDL)原理从给定示例中学习贝叶斯信念网络(BBN)的问题。给定示例,基于MDL的过程将为每个网络计算网络描述长度和给定网络示例的描述长度的总和,并找到具有最小值的某个网络。我们提供了一种搜索算法,可以减少基于MDL的过程的计算时间,同时确保找到具有MDL的网络。与穷举搜索问题相比,该算法采用分支定界(B&B)技术,具有较低的计算复杂度。 Alarm数据库的一些经验实验表明,该算法对于中等大小的各种问题相当有效。虽然B&B技术到目前为止尚未应用于基于MDL /贝叶斯的过程,但是该结果被认为表明了基于MDL的过程优于Cooper和Herskovits提出的贝叶斯过程的优势。

著录项

  • 来源
    《Machine learning》|1996年|462-470|共9页
  • 会议地点 Bari(IT);Bari(IT)
  • 作者

    Joe Suzuki;

  • 作者单位

    Information Systems Laboratory, Stanford University Stanford, CA 94305-4055;

  • 会议组织
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 计算机的应用;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号