| 摘 要: 针对大数据处理中存在的优化问题，提出一种基于凸和非凸优化低秩矩估计的大数据处理算法，并利用 物联网采集与感知的大数据进行算法的实验验证与对比分析。在简单描述凸优化、非凸优化和低秩矩阵优化基础上，设 计了基于凸和非凸优化低秩矩估计的大数据处理算法，并对算法收敛性进行了分析；然后利用物联网感知设备，进行数 据感知与采集，然后利用所设计的算法进行对比实验。通过实验表明，本文算法在训练时和测试时，在归一化平均绝对 误差和运行时间上，具有较好的结果。 |
| 关键词: 凸优化；非凸优化；低秩矩阵；估计；大数据处理 |
|
中图分类号: TP309
文献标识码: A
|
| 基金项目: 2019年度四川水利职业技术学院院级科研项目(KY2019-04)；四川省教育厅自然科学基金一般项目(No.18ZB0498)；四川省水利厅2017年科研计划项目(No.SL2017- 01)的资助. |
|
| Processing Algorithm for Big Data Based on Convex Optimization,Non-convex Optimization and Low-Rank Matrix Estimation |
|
LIU Xiaoxia,WANG Yuning,CHEN Xia
|
( Department of Information Engineering, Sichuan Water Conservancy Vocational College, Chongzhou 611231, China)
|
| Abstract: Aiming to the optimization problem in big data processing,a big data processing algorithm based on convex optimization,non-convex optimization and low rank matrix estimation is proposed,and the experimental verification and comparative analysis of the algorithm are carried out by means of the big data collected and perceived by Internet of Things. On the basis of convex optimization,non-convex optimization and low rank matrix optimization,the big data processing algorithm is designed and the convergence of the algorithm is analyzed.Then data perception and collection are performed by using the Internet of Things perception devices and the proposed algorithm is used to conduct comparative experiments. Experiments show that the proposed algorithm has good results in normalized mean absolute error and running time during training and testing. |
| Keywords: convex optimization;non-convex optimization;low rank matrix;estimation;big data processing |
