The dominant human rights views in a country or region is closely related to the economic and social development of that country or region and is closely related to human rights practices. Since the launch of reform a...The dominant human rights views in a country or region is closely related to the economic and social development of that country or region and is closely related to human rights practices. Since the launch of reform and opening-up, the development of China’s conception of human rights has gone through many stages. The changes of contemporary Chinese human rights conceptions have gone through the stages from the views of human rights to the theories of human rights, to the norms and practices of human rights, and human rights conceptions developed from avoidance to controversy, from passiveness to initiative, and from learning to construction, which reflects the continuous development of China’s human rights cause. By analyzing the journal records through a search of the human rights and its related keywords, it is possible to chart the evolution and historical context of contemporary Chinese human rights theory.展开更多
在众多现实场景中数据以流的形式持续产生,由于流数据具有动态变化的特点,在生成过程中可能产生新的类别,也被称为概念演化.概念演化是流数据挖掘模型预测性能衰退甚至预测失效的主要原因.因此,能及时发现类空间变化并提醒模型做出适应...在众多现实场景中数据以流的形式持续产生,由于流数据具有动态变化的特点,在生成过程中可能产生新的类别,也被称为概念演化.概念演化是流数据挖掘模型预测性能衰退甚至预测失效的主要原因.因此,能及时发现类空间变化并提醒模型做出适应性调节的概念演化检测方法受到广泛关注.然而,目前多数概念演化检测方法基于特征空间静态不变的假设构建算法.在现实场景中,特征空间同样具有动态性,属于开放的空间.具体来说,随时间推移可能出现部分特征消失和新特征产生的现象,从而破坏上述假设并导致已有算法失效.针对这一问题,本文提出一种面向开放特征空间的概念演化检测方法(Concept evolution Detection method for Open Feature space,CD_OF).该方法通过构建微簇集成模型对新进实例分类,对于开放特征空间中的旧特征消失问题,通过利用转移矩阵将旧特征中包含的信息转换到共享特征中;对于新出现的特征,拓展共享特征空间并重构集成模型.在此基础上,根据样本的共享邻域信息定义样本间相似度以检测概念演化,并建立动态衰减模型,以解决开放特征空间下的类消失和类循环问题.实验结果表明,本文所提出的方法能够对开放特征空间中特征的变化作出及时响应,增强概念演化检测的能力,在特征空间变化的真实流数据中与现有方法相比,错误率降低了1.7%~11.4%.展开更多
基金the phased achievement of the major project “Construction and Expression of China’s Human Rights Discourse System” of the National Human Rights Education and Training Base
文摘The dominant human rights views in a country or region is closely related to the economic and social development of that country or region and is closely related to human rights practices. Since the launch of reform and opening-up, the development of China’s conception of human rights has gone through many stages. The changes of contemporary Chinese human rights conceptions have gone through the stages from the views of human rights to the theories of human rights, to the norms and practices of human rights, and human rights conceptions developed from avoidance to controversy, from passiveness to initiative, and from learning to construction, which reflects the continuous development of China’s human rights cause. By analyzing the journal records through a search of the human rights and its related keywords, it is possible to chart the evolution and historical context of contemporary Chinese human rights theory.
文摘在众多现实场景中数据以流的形式持续产生,由于流数据具有动态变化的特点,在生成过程中可能产生新的类别,也被称为概念演化.概念演化是流数据挖掘模型预测性能衰退甚至预测失效的主要原因.因此,能及时发现类空间变化并提醒模型做出适应性调节的概念演化检测方法受到广泛关注.然而,目前多数概念演化检测方法基于特征空间静态不变的假设构建算法.在现实场景中,特征空间同样具有动态性,属于开放的空间.具体来说,随时间推移可能出现部分特征消失和新特征产生的现象,从而破坏上述假设并导致已有算法失效.针对这一问题,本文提出一种面向开放特征空间的概念演化检测方法(Concept evolution Detection method for Open Feature space,CD_OF).该方法通过构建微簇集成模型对新进实例分类,对于开放特征空间中的旧特征消失问题,通过利用转移矩阵将旧特征中包含的信息转换到共享特征中;对于新出现的特征,拓展共享特征空间并重构集成模型.在此基础上,根据样本的共享邻域信息定义样本间相似度以检测概念演化,并建立动态衰减模型,以解决开放特征空间下的类消失和类循环问题.实验结果表明,本文所提出的方法能够对开放特征空间中特征的变化作出及时响应,增强概念演化检测的能力,在特征空间变化的真实流数据中与现有方法相比,错误率降低了1.7%~11.4%.
