The sharp increase of the amount of Internet Chinese text data has significantly prolonged the processing time of classification on these data.In order to solve this problem,this paper proposes and implements a parall...The sharp increase of the amount of Internet Chinese text data has significantly prolonged the processing time of classification on these data.In order to solve this problem,this paper proposes and implements a parallel naive Bayes algorithm(PNBA)for Chinese text classification based on Spark,a parallel memory computing platform for big data.This algorithm has implemented parallel operation throughout the entire training and prediction process of naive Bayes classifier mainly by adopting the programming model of resilient distributed datasets(RDD).For comparison,a PNBA based on Hadoop is also implemented.The test results show that in the same computing environment and for the same text sets,the Spark PNBA is obviously superior to the Hadoop PNBA in terms of key indicators such as speedup ratio and scalability.Therefore,Spark-based parallel algorithms can better meet the requirement of large-scale Chinese text data mining.展开更多
The method of establishing data structures plays an important role in the efficiency of parallel multilevel fast multipole algorithm（PMLFMA）.Considering the main complements of multilevel fast multipole algorithm（M...The method of establishing data structures plays an important role in the efficiency of parallel multilevel fast multipole algorithm（PMLFMA）.Considering the main complements of multilevel fast multipole algorithm（MLFMA） memory,a new parallelization strategy and a modified data octree construction scheme are proposed to further reduce communication in order to improve parallel efficiency.For far interaction,a new scheme called dynamic memory allocation is developed.To analyze the workload balancing performance of a parallel implementation,the original concept of workload balancing factor is introduced and verified by numerical examples.Numerical results show that the above measures improve the parallel efficiency and are suitable for the analysis of electrical large-scale scattering objects.展开更多
提出并实现了一种新的蚁群优化(ACO)并行化策略SHOP(Sharing one pheromone matrix)．主要思想是基于多蚁群在解的构造过程和信息素更新过程中共享同一个信息素矩阵．以ACS和MMAS的SHOP并行实现为例,简要描述了SHOP设计思想和实现过程,...提出并实现了一种新的蚁群优化(ACO)并行化策略SHOP(Sharing one pheromone matrix)．主要思想是基于多蚁群在解的构造过程和信息素更新过程中共享同一个信息素矩阵．以ACS和MMAS的SHOP并行实现为例,简要描述了SHOP设计思想和实现过程,尝试了ACS和MMAS并行混合．以对称TSP测试集为对象,将SHOP的实现与相应串行算法在相同计算环境下的实验结果比较,以及与现有的并行实现进行比较,结果表明SHOP并行策略相对于串行ACO及现有的并行策略具有一定的优势．展开更多
基金Project(KC18071)supported by the Application Foundation Research Program of Xuzhou,ChinaProjects(2017YFC0804401,2017YFC0804409)supported by the National Key R&D Program of China
文摘The sharp increase of the amount of Internet Chinese text data has significantly prolonged the processing time of classification on these data.In order to solve this problem,this paper proposes and implements a parallel naive Bayes algorithm(PNBA)for Chinese text classification based on Spark,a parallel memory computing platform for big data.This algorithm has implemented parallel operation throughout the entire training and prediction process of naive Bayes classifier mainly by adopting the programming model of resilient distributed datasets(RDD).For comparison,a PNBA based on Hadoop is also implemented.The test results show that in the same computing environment and for the same text sets,the Spark PNBA is obviously superior to the Hadoop PNBA in terms of key indicators such as speedup ratio and scalability.Therefore,Spark-based parallel algorithms can better meet the requirement of large-scale Chinese text data mining.
基金supported by the National Basic Research Program of China (973 Program) (61320)
文摘The method of establishing data structures plays an important role in the efficiency of parallel multilevel fast multipole algorithm（PMLFMA）.Considering the main complements of multilevel fast multipole algorithm（MLFMA） memory,a new parallelization strategy and a modified data octree construction scheme are proposed to further reduce communication in order to improve parallel efficiency.For far interaction,a new scheme called dynamic memory allocation is developed.To analyze the workload balancing performance of a parallel implementation,the original concept of workload balancing factor is introduced and verified by numerical examples.Numerical results show that the above measures improve the parallel efficiency and are suitable for the analysis of electrical large-scale scattering objects.
文摘提出并实现了一种新的蚁群优化(ACO)并行化策略SHOP(Sharing one pheromone matrix)．主要思想是基于多蚁群在解的构造过程和信息素更新过程中共享同一个信息素矩阵．以ACS和MMAS的SHOP并行实现为例,简要描述了SHOP设计思想和实现过程,尝试了ACS和MMAS并行混合．以对称TSP测试集为对象,将SHOP的实现与相应串行算法在相同计算环境下的实验结果比较,以及与现有的并行实现进行比较,结果表明SHOP并行策略相对于串行ACO及现有的并行策略具有一定的优势．
