期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
基于Spark的并行KMeans聚类模型研究 被引量:8
1
作者 侯敬儒 吴晟 李英娜 《计算机与数字工程》 2018年第3期537-540,555,共5页
文章基于Spark分布式计算框架设计并实现了并行KMeans聚类模型,并通过该模型在不同规模的Movie Lens数据集上进行训练比对实验,结果表明,该并行KMeans聚类模型适合运行在分布式集群环境下,且并行化计算效率也有不俗的表现;其次通过repar... 文章基于Spark分布式计算框架设计并实现了并行KMeans聚类模型,并通过该模型在不同规模的Movie Lens数据集上进行训练比对实验,结果表明,该并行KMeans聚类模型适合运行在分布式集群环境下,且并行化计算效率也有不俗的表现;其次通过repartition算子设计分片加载数据,优化并行方案,有效减少了模型的训练时间。 展开更多
关键词 SPARK KMeans MovieLens 并行聚类 repartition
下载PDF
Finite Element Methods Applied to the Tubular Linear Stepping Motor
2
作者 Walid El Fezzani Abdessattar Ben Amor 《Journal of Electromagnetic Analysis and Applications》 2013年第5期219-222,共4页
In this paper proposes a Finite Element Methods analyzing applied to the linear tubular stepping actuator. The linear displacement is modeled by means of a layer of finite elements placed in the air gap. The design of... In this paper proposes a Finite Element Methods analyzing applied to the linear tubular stepping actuator. The linear displacement is modeled by means of a layer of finite elements placed in the air gap. The design of the linear stepper motor for achieving a specific performance requires the choice of appropriate tooth geometry. The magnetic field of the actuator has been analyzed using the finite element method over a current-displacement variation. The magneto static field and electromagnetic force was introduced in order to predict before construction, the inductance values according to the displacement and the currents into the coils. The results were obtained for the magnetic flux density distribution and the electromagnetic force for different positions and current. 展开更多
关键词 FINITE Element Methods Linear TUBULAR STEPPING ACTUATOR MESH repartition
下载PDF
An Efficient Parallel Solution Framework for the Linear Solution of Large Systems on PC Clusters
3
作者 Ozgur Kurc Semih Ozmen 《Tsinghua Science and Technology》 SCIE EI CAS 2008年第S1期65-70,共6页
In this paper, a parallel solution framework for the linear static analysis of large structures on PC clusters is presented. The framework consists of two main steps: data preparation and parallel solution. The parall... In this paper, a parallel solution framework for the linear static analysis of large structures on PC clusters is presented. The framework consists of two main steps: data preparation and parallel solution. The parallel solution is performed by a substructure based method with direct solvers. The aim of the data preparation step is to create the best possible substructures so that the parallel solution time is minimized. An actual structural model was solved utilizing both homogeneous and heterogeneous PC clusters to illustrate the performance and applicability of the presented framework. 展开更多
关键词 SUBSTRUCTURE parallel solution heterogeneous clusters homogeneous clusters workload bal-ancing partitioning repartitioning
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部