Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgrounda...Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgroundand developing status of MPI. Then on this basis it will study and analyze the functions and features of MPI ,summa-rize its insufficiencies and gives some suggestions for modification.展开更多
时间复杂性是基于 EM 框架的贝叶斯网络学习算法应用的一个瓶颈问题.本文首先提出一种并行的参数EM 算法来学习具有缺省数据的贝叶斯网络参数,实验表明该算法可有效降低参数学习的时间复杂性.进而将该算法应用到结构 EM 算法中,提出一...时间复杂性是基于 EM 框架的贝叶斯网络学习算法应用的一个瓶颈问题.本文首先提出一种并行的参数EM 算法来学习具有缺省数据的贝叶斯网络参数,实验表明该算法可有效降低参数学习的时间复杂性.进而将该算法应用到结构 EM 算法中,提出一种并行的结构 EM 算法(PL-SEM),PL-SEM 算法并行地计算各个样本的期望充分因子和贝叶斯网络的参数,降低结构学习的时间复杂性.展开更多
文摘Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgroundand developing status of MPI. Then on this basis it will study and analyze the functions and features of MPI ,summa-rize its insufficiencies and gives some suggestions for modification.
文摘时间复杂性是基于 EM 框架的贝叶斯网络学习算法应用的一个瓶颈问题.本文首先提出一种并行的参数EM 算法来学习具有缺省数据的贝叶斯网络参数,实验表明该算法可有效降低参数学习的时间复杂性.进而将该算法应用到结构 EM 算法中,提出一种并行的结构 EM 算法(PL-SEM),PL-SEM 算法并行地计算各个样本的期望充分因子和贝叶斯网络的参数,降低结构学习的时间复杂性.
