In the paper,we define weakly δ-continuous correspondences on super-space,On the basis of δ-open(closed) sets,θ-open(closed) sets and regular open(closed) sets in topological space,some equivalent conditions of thi...In the paper,we define weakly δ-continuous correspondences on super-space,On the basis of δ-open(closed) sets,θ-open(closed) sets and regular open(closed) sets in topological space,some equivalent conditions of this kind of correspondences are obtained,and some applications of subset nets and convergence nets are given.展开更多
弱标记学习是多标记学习的一个重要分支,近几年已被广泛研究并被应用于多标记样本的缺失标记补全和预测等问题.然而,针对特征集合较大、更容易拥有多个语义标记和出现标记缺失的高维数据问题,现有弱标记学习方法普遍易受这类数据包含的...弱标记学习是多标记学习的一个重要分支,近几年已被广泛研究并被应用于多标记样本的缺失标记补全和预测等问题.然而,针对特征集合较大、更容易拥有多个语义标记和出现标记缺失的高维数据问题,现有弱标记学习方法普遍易受这类数据包含的噪声和冗余特征的干扰.为了对高维多标记数据进行准确的分类,提出了一种基于标记与特征依赖最大化的弱标记集成分类方法 En WL.En WL首先在高维数据的特征空间多次利用近邻传播聚类方法,每次选择聚类中心构成具有代表性的特征子集,降低噪声和冗余特征的干扰;再在每个特征子集上训练一个基于标记与特征依赖最大化的半监督多标记分类器;最后,通过投票集成这些分类器实现多标记分类.在多种高维数据集上的实验结果表明,En WL在多种评价度量上的预测性能均优于已有相关方法.展开更多
Ⅰ. PROBLEM AND ASSUMPTIONS In 1978 J. Henry first discussed the globe controllability problem on a semilinear parabolic system with a distributed control. The author obtained the approximate control lability theory o...Ⅰ. PROBLEM AND ASSUMPTIONS In 1978 J. Henry first discussed the globe controllability problem on a semilinear parabolic system with a distributed control. The author obtained the approximate control lability theory on a semilinear abstract control system unifying finited-dimen-展开更多
文摘In the paper,we define weakly δ-continuous correspondences on super-space,On the basis of δ-open(closed) sets,θ-open(closed) sets and regular open(closed) sets in topological space,some equivalent conditions of this kind of correspondences are obtained,and some applications of subset nets and convergence nets are given.
文摘弱标记学习是多标记学习的一个重要分支,近几年已被广泛研究并被应用于多标记样本的缺失标记补全和预测等问题.然而,针对特征集合较大、更容易拥有多个语义标记和出现标记缺失的高维数据问题,现有弱标记学习方法普遍易受这类数据包含的噪声和冗余特征的干扰.为了对高维多标记数据进行准确的分类,提出了一种基于标记与特征依赖最大化的弱标记集成分类方法 En WL.En WL首先在高维数据的特征空间多次利用近邻传播聚类方法,每次选择聚类中心构成具有代表性的特征子集,降低噪声和冗余特征的干扰;再在每个特征子集上训练一个基于标记与特征依赖最大化的半监督多标记分类器;最后,通过投票集成这些分类器实现多标记分类.在多种高维数据集上的实验结果表明,En WL在多种评价度量上的预测性能均优于已有相关方法.
文摘Ⅰ. PROBLEM AND ASSUMPTIONS In 1978 J. Henry first discussed the globe controllability problem on a semilinear parabolic system with a distributed control. The author obtained the approximate control lability theory on a semilinear abstract control system unifying finited-dimen-