|
|||||
|
|
| 基于局部优化具有连续变量的贝叶斯网络结构学习 | |
| 引用本文： | 张剑飞,王辉,周颜军,王双成.基于局部优化具有连续变量的贝叶斯网络结构学习[J].东北师大学报(自然科学版),2006,38(1):27-30. |
| 作者姓名： | 张剑飞 王辉 周颜军 王双成 |
| 作者单位： | 1. 齐齐哈尔大学计算机与控制工程学院,黑龙江,齐齐哈尔,161006;东北师范大学计算机学院,吉林,长春,130024 2. 东北师范大学计算机学院,吉林,长春,130024 3. 上海立信会计学院信息科学院,上海,201600 |
| 基金项目： | 吉林省自然科学基金资助项目(20030517-1) |
| 摘 要： | 概述了具有连续变量的贝叶斯网络结构学习存在的主要问题,给出了基于局部优化的具有连续变量的贝叶斯网络结构学习方法.通过构造局部最优回归模式、局部回归模式的条件组合及环路处理,建立了具有连续变量的贝叶斯网络结构,既可以避免复杂的结构打分运算及结构空间搜索,同时又不会出现由于离散化而导致过多的信息丢失及假依赖现象. |
| 关 键 词： | 连续变量 贝叶斯网络 结构学习 回归 |
| 文章编号： | 1000-1832（2006）01-0027-04 |
| 收稿时间： | 2005-09-27 |
| 修稿时间： | 2005年9月27日 |
Learning Bayesian network structure with continuous variables based on local optimization |
|
| ZHANG Jian-fei,WANG Hui,ZHOU Yan-jun,WANG Shuang-cheng.Learning Bayesian network structure with continuous variables based on local optimization[J].Journal of Northeast Normal University (Natural Science Edition),2006,38(1):27-30. | |
| Authors: | ZHANG Jian-fei WANG Hui ZHOU Yan-jun WANG Shuang-cheng |
| Institution: | 1. College of Computer and Control Engineering, Qiqihr University, Qiqihr 161006,China; 2. College of Computer, Northeast Normal University, Changchtm 130024, China; 3. Department of Information Seienee, Shanghai Lixln University of Commerce, Shanghai 201600,China |
| Abstract: | Main problems are presented in learning Bayesian network structure with continuous variables.To solve these problems,a new method of learning Bayesian network structure with continuous variables based on local optimization is developed.In this method,the Bayesian network structure with continuous variables is set up by constructing local optimization regression models,conditional combination of these models and dealing with loop.This method can avoid complicated structure scoring and structure search.At the same time,the case of leading to miss information and false dependency can be got over. |
| Keywords: | continuous variable bayesian network structure learning regression |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |