正在加载图片...
第15卷第1期 智能系统学报 Vol.15 No.1 2020年1月 CAAI Transactions on Intelligent Systems Jan.2020 D0L:10.11992tis.201904037 基于知识距离的粗糙粒结构的评价模型 杨洁2,王国胤2,张清华2 (1,重庆邮电大学计算智能重庆市重点实验室,重庆400065,2.重庆邮电大学计算机科学与技术学院,重庆400065; 3.重庆邮电大学人工智能学院.重庆400065) 摘要:在粒计算理论中,通过不同的粒计算机制可以生成不同的粒结构。在粗糙集中,对于同一个信息表而 言,通过不同的属性添加顺序可以得到由不同的序贯层次结构,即粗糙粒结构。在粗糙粒结构中,不同的属性 获取顺序导致了对不确定性问题求解的不同程度。因此,如何有效评价粗糙粒结构是一个值得研究的问题。 本文将从知识距离的角度研究这个问题。首先,在前期工作所提出的知识距离框架上提出了一种粗糙近似空 间距离,用于度量粗糙近似空间之间差异性。基于提出的知识距离,研究了粗糙粒结构的结构特征。在粗糙粒 结构中,在对不确定性问题进行求解时,本文希望在约束条件下可以利用尽可能少的知识空间使不确定性降低 达到最大化。基于这个思想并利用以上得出的结论,在属性代价约束条件下,引入了一个评价参数,并在此 基础建立了一种粗糙粒结构的评价模型,该方法实现了在属性代价约束条件下选择粗糙粒结构的功能。最后, 通过实例验证了本文提出的模型的有效性。 关键词:粗糙粒结构;知识距离:不确定性度量:评价模型;粒计算;粗糙集;约束条件:不确定性度量 中图分类号:TP311文献标志码:A文章编号:1673-4785(2020)01-0166-09 中文引用格式:杨洁,王国胤,张清华.基于知识距离的粗糙粒结构的评价模型.智能系统学报,2020,15(1):166-174. 英文引用格式:YANG Jie,.WANG Guoyin,.ZHANG Qinghua..Evaluation model of rough granular structure based on knowledge distance[JI.CAAI transactions on intelligent systems,2020,15(1):166-174. Evaluation model of rough granular structure based on knowledge distance YANG Jie2,WANG Guoyin'2,ZHANG Qinghua23 (1.Chongqing Key Laboratory of Computational Intelligence,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;2.School of computer science and technology,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;3.College of Artificial Intelligence,Chongqing University of Posts and Telecommunications,Chongqing 400065, China) Abstract:In the theory of granular computing(GrC),different granular structures are generated by various grain calcu- lation mechanisms.In rough sets,for the same information table,different attribute adding sequence produces different sequential hierarchical structure,namely the rough granular structure.In rough granular structure,various order of attrib- ute acquisition leads to different effects of solving uncertain problems.This leads to an interesting research topic:how to effectively evaluate the rough granular structures.This problem is solved from the perspective of knowledge distance in the paper.Firstly,the knowledge distance mentioned in our previous works is introduced and then a rough approxima- tion space distance(RASD)is proposed to measure the difference between rough approximate space.On the basis of the knowledge distance mentioned above,the characteristic of rough granular structure(RGS)is investigated.In the rough granular structure,when solving uncertain problem,we expect to to maximize the uncertainty reduction as much as pos- sible by using smaller knowledge space.Then,based on this idea and the above conclusions,an evaluation parameter A is introduced under the constraint of attribute cost,and further,an evaluation model of rough granular structure is estab- lished.This achieves a way to select the rough granular structure according to the constraint.Finally,the effectiveness of this method is verified by an example. Keywords:rough granular structure;knowledge distance;uncertainty measure;evaluation model;granular computing; rough sets;constraint condition;uncertainty measure 收稿日期:2019-04-16 基金项目:国家自然科学基金项目(61572091):贵州省教育厅科 早在1997年,Zadeh教授U就提出了粒计算 技人才成长项目(KY(2018)No.318). 通信作者:杨洁.E-mail:530966074@qq.com 是模糊信息粒化、粗糙集理论和区间计算的超DOI: 10.11992/tis.201904037 基于知识距离的粗糙粒结构的评价模型 杨洁1,2,王国胤1,2,3,张清华1,2,3 (1. 重庆邮电大学 计算智能重庆市重点实验室,重庆 400065; 2. 重庆邮电大学 计算机科学与技术学院,重庆 400065; 3. 重庆邮电大学 人工智能学院,重庆 400065) 摘 要:在粒计算理论中,通过不同的粒计算机制可以生成不同的粒结构。在粗糙集中,对于同一个信息表而 言,通过不同的属性添加顺序可以得到由不同的序贯层次结构,即粗糙粒结构。在粗糙粒结构中,不同的属性 获取顺序导致了对不确定性问题求解的不同程度。因此,如何有效评价粗糙粒结构是一个值得研究的问题。 本文将从知识距离的角度研究这个问题。首先,在前期工作所提出的知识距离框架上提出了一种粗糙近似空 间距离,用于度量粗糙近似空间之间差异性。基于提出的知识距离,研究了粗糙粒结构的结构特征。在粗糙粒 结构中,在对不确定性问题进行求解时,本文希望在约束条件下可以利用尽可能少的知识空间使不确定性降低 达到最大化。基于这个思想并利用以上得出的结论,在属性代价约束条件下,引入了一个评价参数 λ,并在此 基础建立了一种粗糙粒结构的评价模型,该方法实现了在属性代价约束条件下选择粗糙粒结构的功能。最后, 通过实例验证了本文提出的模型的有效性。 关键词:粗糙粒结构;知识距离;不确定性度量;评价模型;粒计算;粗糙集;约束条件;不确定性度量 中图分类号:TP311 文献标志码:A 文章编号:1673−4785(2020)01−0166−09 中文引用格式:杨洁, 王国胤, 张清华. 基于知识距离的粗糙粒结构的评价模型 [J]. 智能系统学报, 2020, 15(1): 166–174. 英文引用格式:YANG Jie, WANG Guoyin, ZHANG Qinghua. Evaluation model of rough granular structure based on knowledge distance[J]. CAAI transactions on intelligent systems, 2020, 15(1): 166–174. Evaluation model of rough granular structure based on knowledge distance YANG Jie1,2 ,WANG Guoyin1,2,3 ,ZHANG Qinghua1,2,3 (1. Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, China; 2. School of computer science and technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, China; 3. College of Artificial Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, China) Abstract: In the theory of granular computing (GrC), different granular structures are generated by various grain calcu￾lation mechanisms. In rough sets, for the same information table, different attribute adding sequence produces different sequential hierarchical structure, namely the rough granular structure. In rough granular structure, various order of attrib￾ute acquisition leads to different effects of solving uncertain problems. This leads to an interesting research topic: how to effectively evaluate the rough granular structures. This problem is solved from the perspective of knowledge distance in the paper. Firstly, the knowledge distance mentioned in our previous works is introduced and then a rough approxima￾tion space distance (RASD) is proposed to measure the difference between rough approximate space. On the basis of the knowledge distance mentioned above, the characteristic of rough granular structure (RGS) is investigated. In the rough granular structure, when solving uncertain problem, we expect to to maximize the uncertainty reduction as much as pos￾sible by using smaller knowledge space. Then, based on this idea and the above conclusions, an evaluation parameter λ is introduced under the constraint of attribute cost, and further, an evaluation model of rough granular structure is estab￾lished. This achieves a way to select the rough granular structure according to the constraint. Finally, the effectiveness of this method is verified by an example. Keywords: rough granular structure; knowledge distance; uncertainty measure; evaluation model; granular computing; rough sets; constraint condition; uncertainty measure 早在 1997 年,Zadeh 教授[1] 就提出了粒计算 是模糊信息粒化、粗糙集理论和区间计算的超 收稿日期:2019−04−16. 基金项目:国家自然科学基金项目 (61572091);贵州省教育厅科 技人才成长项目 (KY(2018)No.318). 通信作者:杨洁. E-mail:530966074@qq.com. 第 15 卷第 1 期 智 能 系 统 学 报 Vol.15 No.1 2020 年 1 月 CAAI Transactions on Intelligent Systems Jan. 2020
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有