针对不确定的传感器数据流,在对国外数据流管理原型系统研究的基础上,采用客户机/服务器体系结构,在Red Hat Linux 9.0平台上部分地实现了不确定数据流数据库系统(UCDS)。详细描述了不确定数据流数据库系统的基本定义、系统的体系结构等...针对不确定的传感器数据流,在对国外数据流管理原型系统研究的基础上,采用客户机/服务器体系结构,在Red Hat Linux 9.0平台上部分地实现了不确定数据流数据库系统(UCDS)。详细描述了不确定数据流数据库系统的基本定义、系统的体系结构等,为不确定性数据库的研究做出了有益的探索。展开更多
In data mining, traditional clustering techniques are not so effective as hoped. The large-scale databases, the disagreements in variable’s measurement levels, and the exhausting distance matrix operation, are the th...In data mining, traditional clustering techniques are not so effective as hoped. The large-scale databases, the disagreements in variable’s measurement levels, and the exhausting distance matrix operation, are the three main obstacles.By using roughset theory, we introduced a new clustering approach in large-scale data analysis without using geometry distance measurements.展开更多
随着对实验精确度要求的不断提高,聚类分析中的不确定性数据聚类也越来越受到关注.然而经典的不确定数据聚类通常假设其概率密度函数(PDF)等信息是已知的,而现实过程中,这些指标并没有那么轻易就能获取.考虑到这些情况,可以利用三角模...随着对实验精确度要求的不断提高,聚类分析中的不确定性数据聚类也越来越受到关注.然而经典的不确定数据聚类通常假设其概率密度函数(PDF)等信息是已知的,而现实过程中,这些指标并没有那么轻易就能获取.考虑到这些情况,可以利用三角模糊数来恰当有效地表示多维不确定性数据,并采用基于三角模糊数的低计算复杂度的距离计算方法,结合K-means基础聚类方法形成一种被命名为UTDK-means(Uncertain triangular fuzzy number data K-means)的聚类方法,而它是基于三角模糊数的.实验结果表明:基于三角模糊数的不确定数据聚类是可行的,具有一定的研究价值.展开更多
In this paper,based on probabilistic and cognitive uncertainties,we set up an uncertain database model and presented the uncertain relational algebra ,and proved that the uncertain normal forms are closed under operat...In this paper,based on probabilistic and cognitive uncertainties,we set up an uncertain database model and presented the uncertain relational algebra ,and proved that the uncertain normal forms are closed under operations of the uncertain relational algebra. At last ,we studied the semantics of the uncertain relation.展开更多
文摘In data mining, traditional clustering techniques are not so effective as hoped. The large-scale databases, the disagreements in variable’s measurement levels, and the exhausting distance matrix operation, are the three main obstacles.By using roughset theory, we introduced a new clustering approach in large-scale data analysis without using geometry distance measurements.
文摘随着对实验精确度要求的不断提高,聚类分析中的不确定性数据聚类也越来越受到关注.然而经典的不确定数据聚类通常假设其概率密度函数(PDF)等信息是已知的,而现实过程中,这些指标并没有那么轻易就能获取.考虑到这些情况,可以利用三角模糊数来恰当有效地表示多维不确定性数据,并采用基于三角模糊数的低计算复杂度的距离计算方法,结合K-means基础聚类方法形成一种被命名为UTDK-means(Uncertain triangular fuzzy number data K-means)的聚类方法,而它是基于三角模糊数的.实验结果表明:基于三角模糊数的不确定数据聚类是可行的,具有一定的研究价值.
文摘In this paper,based on probabilistic and cognitive uncertainties,we set up an uncertain database model and presented the uncertain relational algebra ,and proved that the uncertain normal forms are closed under operations of the uncertain relational algebra. At last ,we studied the semantics of the uncertain relation.
