摘要
对传统影响图进行了扩充,提出了带区间参数的影响图(ID)概念,在此基础上,给出了基于遗传算法(GA)的带区间参数影响图的结构学习方法,采用Gibbs采样算法对该类影响图作了近似推理,并应用于具体实例。实验表明,带区间参数的影响图模型适用于求解模糊事件和值不确定事件的概率。
Based on traditional Influence Diagram (ID), ID with interval probability parameters was defined. Then the method for learning ID structures from interval data and the example of approximate inferences were proposed. Experimental results show that our methods are feasible and efficient, and IDs with interval probability parameters are fit for solving fuzzy or uncertain problems.
出处
《计算机应用》
CSCD
北大核心
2008年第B06期156-159,共4页
journal of Computer Applications
基金
教育部"春晖"计划资助项目(Z2005-2-650003)
云南省自然科学基金资助项目(2005F0009Q)
关键词
影响图
区间概率
遗传算法
结构学习
近似推理
Influence Diagram (ID)
interval probability
Genetic Algorithm (GA)
structure learning
approximate inference