摘要
利用Rough集理论处理案例推理问题具有不需要外界信息和先验知识的优点,对案例库中冗余属性进行简化,能够起到优化案例库的作用,同时能够依赖于统计知识提炼规则并形成多个有效的案例索引,在进行案例检索时可针对不同的检索问题选择恰当的索引快速检索到相似的案例,并进行推理得出相应的问题解决方案。最后,以稀土萃取分离生产过程的产品纯度和料液处理量等生产指标的智能优化设定控制为例,验证了该模型的可行性和精确性。
Rough set theory has the unique merit of having no use for the outside information and the priori knowledge when dealing with the Case-based reasoning（CBR）problem,which can simplify the redundancy attribute in the case library to optimize the case database.Many indexes of case library are formed based on statistical knowledge at the same time.It is possible to retrieve case database according to the difference index and draw a conclusion for the different question.Finally,it verifies the feasibility and the accuracy of this model based on the example of the intelligent optimal setting control for the production indexes just as the product purity and the liquid flow control in the rare earth extraction separation process.
出处
《华东交通大学学报》
2012年第2期42-46,共5页
Journal of East China Jiaotong University
基金
国家自然科学基金项目(50474020)
江西省教育厅青年科学基金项目(GJJ11115)
关键词
案例推理
ROUGH集
数据补全
数据离散
属性约简
case-based reasoning
rough set
data completion
data dispersion
attribute reduction
