针对多类不平衡数据分类准确率低的问题,提出一种基于空间扩展的支持向量机学习算法(support vector machine algorithm based on space spreading,SS-SVM)。根据空间扩展原理,在多维欧式空间中通过空间扩展对少类数据进行上采样,使其...针对多类不平衡数据分类准确率低的问题,提出一种基于空间扩展的支持向量机学习算法(support vector machine algorithm based on space spreading,SS-SVM)。根据空间扩展原理,在多维欧式空间中通过空间扩展对少类数据进行上采样,使其处理数据时减少小区块的影响;降低数据不平衡度以优化分类器组;在扩展的数据集上训练SVM分类器。标准数据集上的实验结果表明,与几种经典的算法相比,SS-SVM在多类不平衡数据分类上可获得令人满意的分类结果,对少类数据分类精度要求较高的问题尤为有效。展开更多
Along the rapid urbanization, the housing problem of medium and low-income residents in cities has been one of social problems, which is drawing the attention of government in the world for a long time. In present Chi...Along the rapid urbanization, the housing problem of medium and low-income residents in cities has been one of social problems, which is drawing the attention of government in the world for a long time. In present China, both the national policy and market of housing system are in the process of perfecting, which pay more attention to the low-income residents and rural residents. This paper chooses Tianjin, the third pole of China development as the example for research. Tianjin public housing residential district planning explored a mode of ＂large-scale mix and small-scale pure＂ for the whole homeland development. It confirms that regional coordination, ecological concepts, green transport, space characteristics and appropriate technology is important in planning.展开更多
文摘针对多类不平衡数据分类准确率低的问题,提出一种基于空间扩展的支持向量机学习算法(support vector machine algorithm based on space spreading,SS-SVM)。根据空间扩展原理,在多维欧式空间中通过空间扩展对少类数据进行上采样,使其处理数据时减少小区块的影响;降低数据不平衡度以优化分类器组;在扩展的数据集上训练SVM分类器。标准数据集上的实验结果表明,与几种经典的算法相比,SS-SVM在多类不平衡数据分类上可获得令人满意的分类结果,对少类数据分类精度要求较高的问题尤为有效。
文摘Along the rapid urbanization, the housing problem of medium and low-income residents in cities has been one of social problems, which is drawing the attention of government in the world for a long time. In present China, both the national policy and market of housing system are in the process of perfecting, which pay more attention to the low-income residents and rural residents. This paper chooses Tianjin, the third pole of China development as the example for research. Tianjin public housing residential district planning explored a mode of ＂large-scale mix and small-scale pure＂ for the whole homeland development. It confirms that regional coordination, ecological concepts, green transport, space characteristics and appropriate technology is important in planning.
