Deep network clustering has a large number of applications in the web. It needs to cluster a large number of cloud computing sources to get enough data, such as multimedia data search, group-buying site information aggregation. The success of the application depends on the efficiency and effectiveness of the clustered cloud computing source. The current research focuses on the correlation between clustering and cloud computing sources and ignores the overlapping relationship between cloud computing sources, so that the same result data on different cloud computing sources are repeatedly clustered, increasing the clustering overhead and the workload of cloud computing source. In order to improve the efficiency of deep-network clustering, this paper presents a novel tuple-level full-parameter symplectic matrix method to estimate and use clustering data on cloud computing sources to select high-correlation, low-overlap cloud computing sources. The method is divided into two stages: The offline stage, we face the tuple level to make the cloud computing source full parameter symplectic matrix, obtain the sample data; In the online phase, we iteratively estimate the coverage and overlap rates of clustering on cloud computing sources for sample data, and use a heuristic strategy for efficiently discovering low-overlap cloud computing sources. Experimental results show that the proposed method can significantly improve the accuracy and efficiency of overlapping precision clustering.
Ma, Jinling. Cloud computing accurate clustering algorithm for full parameter symplectic matrix[J].
Boletin Tecnico/Technical Bulletin,2017,55(7):157-164.
APA
Ma, Jinling.(2017).Cloud computing accurate clustering algorithm for full parameter symplectic matrix.Boletin Tecnico/Technical Bulletin,55(7),157-164.
MLA
Ma, Jinling."Cloud computing accurate clustering algorithm for full parameter symplectic matrix".Boletin Tecnico/Technical Bulletin 55.7(2017):157-164.
修改评论