Application and twice extraction of information based on singular value decomposition
Changsen Yuan, Jiamei Wang, Jing Fan, Rui Lin · 2016
With the popularization and application of computer technology, digital information filled every corner of human society. Artificial management of digital information has been unable to adapt to the development of society, therefore, efficient management and accurate positioning technology of the mass of information has become research hot spots of many research groups. The work proposed in this paper is application and twice extraction of information basing on singular value decomposition. This approach will use the singular value decomposition for the train matrix, modify the eigenvalue by comparison algorithm, generate new train matrix, and then use the principal component analysis of secondary data processing. Our approach not only reduces redundant information and dimension of the original data, but it eliminates the relevance of each input variable. Using extensive simulations and a number of experiments involving proportional data, we show the merits and the accuracy of the proposed approach.