A Belief Rule Extraction Method Based on Rough Sets

2018 2nd International Conference on Artificial Intelligence Applications and Technologies

计算机科学

Yu, Meng^1 ; Huang, Jian^1 ; Kong, Jiangtao^1

National University of Defense Technology, Changsha, China^1

关键词: Belief rule base;    Belief rules;    Data set;    Equivalent class;    Expert experience;    Parameter optimization problems;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/435/1/012053/pdf
DOI  :  10.1088/1757-899X/435/1/012053

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Most of the existing research on belief rule-base focused on parameter optimization problem. However, there are few studies on the problem of establishing the initial belief rule-base due to the difficulty in obtaining expert experience and the unconformity of standards. Based on this, a method of extracting belief rule based on rough sets is proposed. First, the initial attribute index is used to divide the data set into equivalent classes. Then, the belief rule is extracted with the knowledge of rough sets. Finally, the extracted belief rule is used for reasoning and the index is optimized according to the result of reasoning. An example of pipeline leak detection is introduced in the experiment section. Experimental results show that the method is effective and feasible.