文章导读
总览 评价 刘强 1, , 刘波 1,* , 康同曦 1, , 肖燕珊 2, ( 1、 广东工业大学自动化学院; 2、 广东工业大学计算机学院; ) 摘要: 在数据挖掘研究领域,特征选择已经成为一个重要的研究课题,这是因为现实的数据集常常含有高维的特征,尽管这可以使
刘强1,, 刘波1,*, 康同曦1,, 肖燕珊2,
(
1、广东工业大学自动化学院; 2、广东工业大学计算机学院; )
摘要:
在数据挖掘研究领域,特征选择已经成为一个重要的研究课题,这是因为现实的数据集常常含有高维的特征,尽管这可以使信息更加充分,但对分类器的设计也提出了更高的要求。随着特征维数的增加,特征中的不相关信息和冗余信息也会相应的增多。特征选择的作用就是使数据降维,剔除不相关特征和冗余特征,减少建立模型的训练时间,提高分类的准确率。通过学习了一种基于全局最小冗余的特征选择算法并将其应用到多视角分类中,在实验中与传统的多视角分类算法比较,具有更高的分类准确率。
关键词:
特征选择;最小冗余;多视角学习
LIU Qiang1,, LIU Bo1,*, KANG Tongxi1,, XIAO Yanshan2,
(
1、School of Automation, Guangdong University of Technology; 2、School of Computers, Guangdong University of Technology; )
Abstract:
Feature selection has been an important research topic in data mining, because the real data sets often have high-dimensional features, although the information can be more fully, but the design of classifier were put forward to higher requirements. With the increase of the dimension, the irrelevant information and redundant information will also increased. The effect of feature selection is to reduce the data dimension, excluding irrelevant feature and redundant feature and reduce the training time of building model, improving the classification accuracy. By studied a new feature selection algorithm based on global redundancy minimization and apply it to multi-view classification, compared it with traditional multi-view classification algorithm in the experiments has shown higher classification accuracy.?????
Tag:
点此返回栏目查看更多>>>参考论文