摘要
随着大数据时代的到来,高维数组的分析越来越困难。介绍了两种大数据线性降维的方法,通过构造计算矩阵,求解矩阵的特征值和特征向量,可以非常容易地发现数据的极小维度表示,降低后期大数据分析的复杂度。通过采用这些方法进行数据降维,可以找出数据内在的相互关系,非常有利于数据分布规律的分析。
With the large volume of data coming,high dimensional data is very difficult to be analyzed.The two ways are introduced to reduce the dimensionality of large volumes of data.These methods are able to reveal low dimensional structure of high dimensional data from the top or bottom eigenvectors of specially constructed matrices,and can reduce the complexity of data analysis.Based on these ways,the intrinsic connection between the data can be got easily,and which is helpful to research the distributing rules.
出处
《航天电子对抗》
2014年第4期58-60,共3页
Aerospace Electronic Warfare
关键词
降维
主分量分析
多维尺度分析
dimensionality reduction
principal component analysis
metric multidimensional scaling
