为了提高数据流频繁模式挖掘的效率,文中基于经典的数据流频繁模式挖掘算法FP-Stream和分布式并行计算原理,设计了一种分布式并行化数据流频繁模式挖掘算法—DPFP-Stream(Distributed Parallel Algorithm of Mining Frequent Pattern on...为了提高数据流频繁模式挖掘的效率,文中基于经典的数据流频繁模式挖掘算法FP-Stream和分布式并行计算原理,设计了一种分布式并行化数据流频繁模式挖掘算法—DPFP-Stream(Distributed Parallel Algorithm of Mining Frequent Pattern on Data Stream)。该算法将建立频繁模式树的任务分为local和global两部分,并设置了参数"当前时间";将到达的流数据平均分配到多个不同的local节点,各local节点使用FP-Growth算法产生该单位时间内本节点的候选频繁项集,并按照单位时间将候选频繁项集及其支持度计数打包发送至global节点;global节点按"当前时间"合并各local节点的中间结果并更新模式树Pattern-Tree。在分布式数据流计算平台Storm上进行的算法实现和性能测试结果表明,DPFP-Stream算法的计算效率能够随着local节点或local bolt线程的增加而提高,适用于高效挖掘数据流中的频繁模式。展开更多
传统的主曲线算法在小规模数据集上能获得良好的效果,但单节点的计算和存储能力都不能满足海量数据主曲线的提取要求,而算法分布式并行化是目前解决该类问题最有效的途径之一。本文提出基于MapReduce框架的分布式软K段主曲线算法(Distri...传统的主曲线算法在小规模数据集上能获得良好的效果,但单节点的计算和存储能力都不能满足海量数据主曲线的提取要求,而算法分布式并行化是目前解决该类问题最有效的途径之一。本文提出基于MapReduce框架的分布式软K段主曲线算法(Distributed soft k-segments principal curve,DisSKPC)。首先,基于分布式K-Means算法,采用递归粒化方法对数据集进行粒化,以确定粒的大小并保证粒中数据的关联性。然后调用软K段主曲线算法计算每个粒数据的局部主成分线段,并提出用噪声方差来消除在高密集、高曲率的数据区域可能产生的过拟合线段。最后借助哈密顿路径和贪婪算法连接这些局部主成分线段,形成一条通过数据云中间的最佳曲线。实验结果表明,本文所提出的DisSKPC算法具有良好的可行性和扩展性。展开更多
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 objective of this research is to realize a composite nonlinear feedback control approach for a class of linear and nonlinear systems with parallel-distributed compensation along with sliding mode control technique...The objective of this research is to realize a composite nonlinear feedback control approach for a class of linear and nonlinear systems with parallel-distributed compensation along with sliding mode control technique.The proposed composite nonlinear feedback control approach consists of two parts.In a word,the first part provides the stability of the closed-loop system and the fast convergence response,as long as the second one improves transient response.In this research,the genetic algorithm in line with the fuzzy logic is designed to calculate constant controller coefficients and optimize the control effort.The effectiveness of the proposed design is demonstrated by servo position control system and inverted pendulum system with DC motor simulation results.展开更多
It is shown that the total life of a parallel system with independent and identical (i.i.d.) exponential components is smaller in the right spread order than an exponential life with the sane mean as the system. As ...It is shown that the total life of a parallel system with independent and identical (i.i.d.) exponential components is smaller in the right spread order than an exponential life with the sane mean as the system. As applications, simple upper bounds for the mean and the variance of the life length of a parallel system with i.i.d. NBUE components are established, as well as the preservation property of the convolution of NBUE and exponential randorn variables.展开更多
文摘为了提高数据流频繁模式挖掘的效率,文中基于经典的数据流频繁模式挖掘算法FP-Stream和分布式并行计算原理,设计了一种分布式并行化数据流频繁模式挖掘算法—DPFP-Stream(Distributed Parallel Algorithm of Mining Frequent Pattern on Data Stream)。该算法将建立频繁模式树的任务分为local和global两部分,并设置了参数"当前时间";将到达的流数据平均分配到多个不同的local节点,各local节点使用FP-Growth算法产生该单位时间内本节点的候选频繁项集,并按照单位时间将候选频繁项集及其支持度计数打包发送至global节点;global节点按"当前时间"合并各local节点的中间结果并更新模式树Pattern-Tree。在分布式数据流计算平台Storm上进行的算法实现和性能测试结果表明,DPFP-Stream算法的计算效率能够随着local节点或local bolt线程的增加而提高,适用于高效挖掘数据流中的频繁模式。
文摘传统的主曲线算法在小规模数据集上能获得良好的效果,但单节点的计算和存储能力都不能满足海量数据主曲线的提取要求,而算法分布式并行化是目前解决该类问题最有效的途径之一。本文提出基于MapReduce框架的分布式软K段主曲线算法(Distributed soft k-segments principal curve,DisSKPC)。首先,基于分布式K-Means算法,采用递归粒化方法对数据集进行粒化,以确定粒的大小并保证粒中数据的关联性。然后调用软K段主曲线算法计算每个粒数据的局部主成分线段,并提出用噪声方差来消除在高密集、高曲率的数据区域可能产生的过拟合线段。最后借助哈密顿路径和贪婪算法连接这些局部主成分线段,形成一条通过数据云中间的最佳曲线。实验结果表明,本文所提出的DisSKPC算法具有良好的可行性和扩展性。
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.
文摘The objective of this research is to realize a composite nonlinear feedback control approach for a class of linear and nonlinear systems with parallel-distributed compensation along with sliding mode control technique.The proposed composite nonlinear feedback control approach consists of two parts.In a word,the first part provides the stability of the closed-loop system and the fast convergence response,as long as the second one improves transient response.In this research,the genetic algorithm in line with the fuzzy logic is designed to calculate constant controller coefficients and optimize the control effort.The effectiveness of the proposed design is demonstrated by servo position control system and inverted pendulum system with DC motor simulation results.
基金Supported by National Natural Science Foundation of China under Grant No.10201010Action Programming of Lanzhou University.
文摘It is shown that the total life of a parallel system with independent and identical (i.i.d.) exponential components is smaller in the right spread order than an exponential life with the sane mean as the system. As applications, simple upper bounds for the mean and the variance of the life length of a parallel system with i.i.d. NBUE components are established, as well as the preservation property of the convolution of NBUE and exponential randorn variables.