| 基于奇异值分解的简化数据应用 |
| |
| 引用本文： | 陈卉妍,张仁龙.基于奇异值分解的简化数据应用[J].教育技术导刊,2019,18(8):162-165. |
| |
| 作者姓名： | 陈卉妍 张仁龙 |
| |
| 作者单位： | 北京农学院 计算机与信息工程学院，北京 102206 |
| |
| 摘 要： | 奇异值分解是提取数据特征信息的一种强大工具，其应用可以从信息检索领域扩展到金融、医疗、统计学等各领域，是简化数据、相似度计算的一种有效方法。对奇异值分解原理和特性进行阐述，介绍了基于Python与其相关科学计算库的奇异值分解过程和相似度算法，解释了将庞大的数据矩阵映射到低维空间的转换过程，图像数据通过奇异值分解较原始图像压缩了近8倍。分别对SVD在推荐系统和图像压缩两方面的具体应用进行描述，总结出奇异值分解在数据降维中的强大应用和良好前景。
|
| 关 键 词： | 奇异值分解 推荐引擎 压缩图像 相似度计算 数据降维 |
| 收稿时间： | 2019-06-03 |
Simplified Data Application Based on Singular Value Decomposition |
| |
| CHEN Hui-yan,ZHANG Ren-long.Simplified Data Application Based on Singular Value Decomposition[J].Introduction of Educational Technology,2019,18(8):162-165. |
| |
| Authors: | CHEN Hui-yan ZHANG Ren-long |
| |
| Institution: | College of Computer and Information Engineering， Beijing University of Agriculture， Beijing 102206， China |
| |
| Abstract: | Singular value decomposition is a powerful tool to extract data feature information. Its application can be extended from information retrieval to finance， medical， statistics and other fields. It is an effective method to simplify data and similarity calculation. The principle and characteristics of singular value decomposition are expounded in this paper. The process and similarity algorithm of singular value decomposition based on Python and its related scientific computing library are introduced. The mapping of huge data matrix to low-dimensional space is explained. The image data is nearly 8 times compressed by the singular value decomposition than the original image. The specific applications of SVD in recommendation system and image compression are described respectively， and the powerful application and broad prospects of singular value decomposition in data dimension reduction are summarized. |
| |
| Keywords: | Singular value decomposition recommendation engine compressed image similarity calculation data dimensionality reduction |
|
| 点击此处可从《教育技术导刊》浏览原始摘要信息 |
| 点击此处可从《教育技术导刊》下载免费的PDF全文 |
