| 基于集群的集中调度式RSOM并行训练方法 |
| |
| 引用本文： | 夏胜平,刘建军,袁振涛,虞华,张乐锋,郁文贤.基于集群的集中调度式RSOM并行训练方法[J].系统工程与电子技术,2007,29(3):459-463. |
| |
| 作者姓名： | 夏胜平 刘建军 袁振涛 虞华 张乐锋 郁文贤 |
| |
| 作者单位： | 国防科技大学电子科学与工程学院ATR重点实验室,湖南,长沙,410073 |
| |
| 摘 要： | 聚类算法常需处理海量和高维的大数据集,为了提高算法的效率,设计并行的聚类算法,以提供更好的计算能力是十分必要的。集群系统是目前流行的高效并行计算环境,针对大数据集聚类问题,集群系统的运用是一种很好的选择。由于RSOM(recursive self-organizing mapping,RSOM)模型具有内在的层次化及分布式结构特点,本文研究了基于集群的集中调度式RSOM树并行算法,并通过视频图像特征集实例证实了算法的可行性。
|
| 关 键 词： | 数据聚类 并行计算 集群系统 视频图像 |
| 文章编号： | 1001-506X(2007)03-0459-05 |
| 修稿时间： | 2006年2月16日 |
Center-controlled RSOM parallel training algorithm for cluster-system |
| |
| XIA Sheng-ping,LIU Jian-jun,YUAN Zhen-tao,YU Hua,ZHANG Le-feng,YU Wen-xian.Center-controlled RSOM parallel training algorithm for cluster-system[J].System Engineering and Electronics,2007,29(3):459-463. |
| |
| Authors: | XIA Sheng-ping LIU Jian-jun YUAN Zhen-tao YU Hua ZHANG Le-feng YU Wen-xian |
| |
| Abstract: | In order to improve the computing-efficiency of data clustering for large data sets with high dimensionality,it is necessary to research a good parallel clustering algorithm.Cluster-system,as a new and good environment for high-efficiency parallel computing,is a good choice.Considering the hierarchical and distributed structure of a RSOM tree,the nature for parallel computing,a Clustersystem based and center-controlled RSOM parallel training algorithm is proposed.The performance of this method is tested with the large feature data sets which are extracted from a large amount of video pictures. |
| |
| Keywords: | RSOM |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
