正在加载图片...
第12卷第3期 智能系统学报 Vol.12 No.3 2017年6月 CAAI Transactions on Intelligent Systems Jun.2017 D0I:10.11992/is.21704025 网络出版地址:http:/kns.cmki.net/kcms/detail/23.1538.TP.20170703.1853.010.html 广义分布保持属性约简研究 高学义12,张楠2,童向荣12,姜丽丽12 (1.烟台大学数据科学与智能技术山东省高校重,点实验室,山东烟台264005:2.烟台大学计算机与控制工程学 院,山东烟台264005) 摘要:属性约简是粗糙集理论的重要研究内容之一。分布约简保证约简前后每个对象的概率分布保持不变,即保 证每条规则的置信度在约简前后不发生改变。实际应用中,人们往往更加关注可信度较高或较低的规则。因此,在 本文中引入了广义分布保持属性约简,该属性约简可以保证规则的置信度P(P∈[0,α]或[B,1])在约简前后不变。 同时,给出了广义分布保持属性约简的判定方法与基于差别矩阵的广义分布保持属性约简算法,深入讨论了几种特 殊情形下的广义分布保持约简。最后,在4个UCI数据集上进行的实验分析表明,几种特殊情形下的广义分布保持 属性约简可退化为已有的一些属性约简,且在不同置信区间下求得的广义分布保持属性约简存在包含关系,验证了 相关结论的正确性。 关键词:分布保持:属性约简:粗糙集:概率分布:差别矩阵 中图分类号:TP181文献标志码:A文章编号:1673-4785(2017)03-0377-09 中文引用格式:高学义,张楠,童向荣,等.广义分布保持属性约简研究[J].智能系统学报,2017,12(3):377-385 英文引用格式:GAO Xueyi,ZHANG Nan,TONG Xiangrong,etat.Research on attribute reduction using generalized distribution preservation[J].CAAI transactions on intelligent systems,2017,12(3):377-385. Research on attribute reduction using generalized distribution preservation GAO Xueyi2,ZHANG Nan'2,TONG Xiangrong'2,JIANG Lili2 (1.Key Lab for Data Science and Intelligent Technology of Shandong Higher Education Institutes,Yantai University,Yantai 264005, China;2.School of Computer and Control Engineering,Yantai University,Yantai 264005,China) Abstract:Attribute reduction is a pertinent issue in rough set theory.Distribution reduction ensures that the probability distribution of each target does not change before and after reduction;i.e.,it ensures that the confidence of every rule remains unchanged before and after reduction.In actual applications,people are often interested in rules that have higher or lower confidences.Thus,attribute reduction based on generalized distribution preservation is proposed in this paper.Confidences in [0,a]or [B,1]were unchanged using the proposed technique.We also propose judgment methods for generalized-distribution-preservation attribute reduction and investigate the generalized attribute-reduction algorithm based on a discernibility matrix.Some special cases with respect to generalized-distribution-preservation attribute reduction are discussed in depth.Finally,experiments on four data sets downloaded from UCI show that some special cases with respect to generalized distribution preservation reduction could degenerate into some existing attribute reductions and inclusion relations exist in generalized distribution preservation attribute reduction under different confidence intervals,verifying the correctness of the relevant conclusions. Keywords:distribution preservation;attribute reduction;rough sets;probability distribution;discernibility matrix 粗糙集理论是由波兰学者Pawlak教授于1982 年提出的一种用于处理和分析不确定、不精确数据 的数学方法与工具[1-]。目前,粗糙集理论在机器 收稿日期:2017-04-19.网络出版日期:2017-07-03. 基金项目:国家自然科学基金项目(61403329,61572418,61502410. 学习、决策分析、模式识别、数据挖掘和智能信息处 61572419):山东省自然科学基金项目(ZR2013FQ020, 理等领域得到了广泛应用。 ZR2015PF010):山东省高等学校科技计划项目(J15LN09 116LN17). 属性约简或知识约简是粗糙集理论的重要研 通信作者:张楠.E-mail:zhangnane0851@163.com.第 12 卷第 3 期 智 能 系 统 学 报 Vol.12 №.3 2017 年 6 月 CAAI Transactions on Intelligent Systems Jun. 2017 DOI:10.11992 / tis. 21704025 网络出版地址:http: / / kns.cnki.net / kcms/ detail / 23.1538.TP.20170703.1853.010.html 广义分布保持属性约简研究 高学义1,2 ,张楠1,2 ,童向荣1,2 ,姜丽丽1,2 (1. 烟台大学 数据科学与智能技术山东省高校重点实验室,山东 烟台 264005; 2. 烟台大学 计算机与控制工程学 院,山东 烟台 264005) 摘 要:属性约简是粗糙集理论的重要研究内容之一。 分布约简保证约简前后每个对象的概率分布保持不变,即保 证每条规则的置信度在约简前后不发生改变。 实际应用中,人们往往更加关注可信度较高或较低的规则。 因此,在 本文中引入了广义分布保持属性约简,该属性约简可以保证规则的置信度 P(P∈[0,α]或[ β,1])在约简前后不变。 同时,给出了广义分布保持属性约简的判定方法与基于差别矩阵的广义分布保持属性约简算法,深入讨论了几种特 殊情形下的广义分布保持约简。 最后,在 4 个 UCI 数据集上进行的实验分析表明,几种特殊情形下的广义分布保持 属性约简可退化为已有的一些属性约简,且在不同置信区间下求得的广义分布保持属性约简存在包含关系,验证了 相关结论的正确性。 关键词:分布保持;属性约简;粗糙集;概率分布;差别矩阵 中图分类号:TP181 文献标志码:A 文章编号:1673-4785(2017)03-0377-09 中文引用格式:高学义,张楠,童向荣,等.广义分布保持属性约简研究[J]. 智能系统学报, 2017, 12(3): 377-385. 英文引用格式:GAO Xueyi,ZHANG Nan,TONG Xiangrong,et at. Research on attribute reduction using generalized distribution preservation[J]. CAAI transactions on intelligent systems, 2017, 12(3): 377-385. Research on attribute reduction using generalized distribution preservation GAO Xueyi 1,2 , ZHANG Nan 1,2 , TONG Xiangrong 1,2 , JIANG Lili 1,2 (1.Key Lab for Data Science and Intelligent Technology of Shandong Higher Education Institutes, Yantai University, Yantai 264005, China; 2. School of Computer and Control Engineering, Yantai University, Yantai 264005, China) Abstract:Attribute reduction is a pertinent issue in rough set theory. Distribution reduction ensures that the probability distribution of each target does not change before and after reduction; i.e., it ensures that the confidence of every rule remains unchanged before and after reduction. In actual applications, people are often interested in rules that have higher or lower confidences. Thus, attribute reduction based on generalized distribution preservation is proposed in this paper. Confidences in [0, α] or [β, 1] were unchanged using the proposed technique. We also propose judgment methods for generalized⁃distribution⁃preservation attribute reduction and investigate the generalized attribute⁃reduction algorithm based on a discernibility matrix. Some special cases with respect to generalized⁃distribution⁃preservation attribute reduction are discussed in depth. Finally, experiments on four data sets downloaded from UCI show that some special cases with respect to generalized distribution preservation reduction could degenerate into some existing attribute reductions and inclusion relations exist in generalized distribution preservation attribute reduction under different confidence intervals, verifying the correctness of the relevant conclusions. Keywords: distribution preservation; attribute reduction; rough sets; probability distribution; discernibility matrix 收稿日期:2017-04-19. 网络出版日期:2017-07-03. 基金项目:国家自然科学基金项目( 61403329, 61572418, 61502410, 61572419);山 东 省 自 然 科 学 基 金 项 目 ( ZR2013FQ020, ZR2015PF 010);山东省高等学校科技计划项目( J15LN09, 116LN17). 通信作者:张楠.E⁃mail:zhangnan0851@ 163.com. 粗糙集理论是由波兰学者 Pawlak 教授于 1982 年提出的一种用于处理和分析不确定、不精确数据 的数学方法与工具[1-4] 。 目前,粗糙集理论在机器 学习、决策分析、模式识别、数据挖掘和智能信息处 理等领域得到了广泛应用。 属性约简或知识约简是粗糙集理论的重要研
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有