新的贝叶斯网络结构学习方法

贝叶斯网络是一种将贝叶斯概率方法和有向无环图的网络拓扑结构有机结合的表示模型,它描述了数据项及数据项之间的非线性依赖关系.报告了贝叶斯网络研究的现状,并针对传统算法需要主观规定网络中结点顺序的缺点,提出了一个新的可以在无约束条件下,根据观测得到的训练样本集的概率关系,自动完成学习贝叶斯网络结构的新方法.     

XFoms客户端处理器的研究与实现

已有的HTML表单已经不能满足当前各类新兴的Web应用的需求.虽然W3C已经提出了新一代Web表单标准--XForms,但XForms处理器的研究尚在起步阶段.阐述了XForms表单的分层定义思想以及处理模型,分析和研究了XForms规范,提供了一个基于本规范的XForms客户端处理器的实现方法.     

网格任务调度方法研究

网格计算中的关键问题之一是计算任务在各个资源之间的调度。提出了基于量子遗传算法（QGA）的网格任务调度算法,以减少调度时间为主要目标,增加资源利用率为次要目标。该算法采用量子比特间接编码的方式,通过有向无环图（DAG）来描述子任务间的依赖关系,根据深度值来给子任务的执行顺序进行排序。仿真结果显示,无论是任务完成时间还是资源利用率,此方法都明显优于基于遗传算法（GA）的网格调度算法。     

基于有向无环图的成本-时间优化调度算法

网格计算环境下,基于有向无环图(DAG)的成本-时间优化调度算法运用经济规律把网格用户的任务映射到网格资源中运行.OGS算法考虑了任务间的优先关系,使得任务完成时间最小,但没考虑到在网格环境中所需的成本.Nimrod/G模型中提出基于时间和成本限制下的优化调度算法(DBC)考虑了时间和成本,但没考虑任务问的优先关系.本文综合考虑了成本-时间因素以及任务间的优先关系,在不增加完成时间的基础上,把任务映射到价格便宜的机器上,提出了基于有向无环图的成本-时间优化调度算法.通过仿真表明,相对OGS算法,该算法减少了所需成本.     

基于动态依赖的类间测试顺序研究

类间集成测试顺序决定着测试成本的大小,为了得到合适的测试顺序,提出了一种基于动态依赖的类间测试顺序的方法。首先分析对象关系图中类间依赖关系,然后运用边删除规则去除环路,最后运用有向无环图的拓扑序列给出类的测试顺序。仿真结果表明,本文的方法较Briand的方法减少了42%的测试桩。此方法满足最小化测试桩的需要,提高了测试效率,减少了测试成本。     

基于关联规则和拓扑序列的分类器链方法

在分类器链方法中, 如何确定标签学习次序至关重要, 为此, 提出一种基于关联规则和拓扑序列的分类器链方法(TSECC). 首先结合频繁模式设计了一种基于强关联规则的标签依赖度量策略; 接下来通过标签间依赖关系构建有向无环图, 对图中所有顶点进行拓扑排序; 最后将得到的拓扑序列作为分类器链方法中标签的学习次序, 对每个标签的分类器依次迭代更新. 特别地, 为减少无标签依赖或标签依赖度较低的“孤独”标签对其余标签预测性能的影响, 将“孤独”标签排在拓扑序列之外, 利用二元关联模型训练. 在多种公共多标签数据集上的实验结果表明, TSECC能够有效提升分类性能.     

贝叶斯网络结构模型的构建

贝叶斯网络结构是一种将贝叶斯概率方法和有向无环图的网络拓扑结构有机结合的表示模型,它描述了数据项及其依赖关系,并根据各个变量之间概率关系建立图论模型,但是如何获取具有丢失数据的网络结构是一个急需解决的问题.本文提出一个基于Kullback-Leibler(KL)散度的贝叶斯网络结构学习的KLBN(Kullback-Leibler Bayesian Network)算法.实验结果表明,KLBN算法在可靠性方面明显优于传统的具有丢失数据的贝叶斯网络结构学习算法.     

基于有向无环图的成本-时间优化调度算法

网格计算环境下,基于有向无环图（DAG）的成本-时间优化调度算法运用经济规律把网格用户的任务映射到网格资源中运行。OGS算法考虑了任务间的优先关系,使得任务完成时间最小,但没考虑到在网格环境中所需的成本。Nimrod/G模型中提出基于时间和成本限制下的优化调度算法（DBC）考虑了时间和成本,但没考虑任务间的优先关系。本文综合考虑了成本-时间因素以及任务间的优先关系,在不增加完成时间的基础上,把任务映射到价格便宜的机器上,提出了基于有向无环图的成本-时间优化调度算法。通过仿真表明,相对OGS算法,该算法减少了所需成本。     

报表系统中公式依赖关系分析及计算性能优化

为提升Web报表系统中公式计算的效率,建立了公式计算性能优化的模型.提出了一种公式间依赖关系分析的方法,自适应构建公式间的依赖关系图;在构建的依赖关系图的基础上,进一步提出了高效的层次化拓扑排序算法,极大的提高了报表中公式计算效率,减小报表系统每张报表的表内公式计算的总执行时间.理论分析和实验结果表明,该模型具有较强的可行性和算法高效性.     

基于Flink的任务调度策略

针对大数据流式计算平台拓扑中因各关键节点上任务间不同类型的通信方式导致的通信开销较大问题,提出一种Flink环境下的任务调度策略。通过各任务间数据流大小确定拓扑边权重,将有向无环图转化为拓扑关键路径模型,在保证关键路径上节点负载差异较小的同时,最小化关键任务的节点间通信开销。实验结果表明,该算法与Flink平台现有的任务调度策略相比,在WordCount和TwitterSentiment作业执行过程中计算平均时延降低了13.09%,有效提升了系统性能。     

Topological ordering algorithm for LDAG

Directed Acyclic Graph (DAG) is an important tool for workflow modeling and data provenance management. In these applications, DAG usually performs well. Yet for some workflow applications, except data or control dependencies between atomic tasks, there exists another requirement that each atomic task should be accomplished at an expected stage. Therefore, this paper proposes an improved DAG model – LDAG, in which each vertex has a level. Three cases of the level of vertices are discussed. For a reasonable one of these cases, this paper proposes a topological ordering algorithm and proves its correctness. In addition, it discusses the complexity of the algorithm and some other relevant problems.     

一种新的基于邻接矩阵的拓扑排序算法

为了降低基于邻接矩阵的拓扑排序算法的复杂性，将单顶点算法框架扩展成集合算法框架，给出一些便于进行拓扑排序的有向无环图的性质。在此基础上，定义了适合进行弧删除操作和无前驱顶点判断的邻接矩阵运算，给出了有向弧邻接矩阵的存储方案，最终提出了一种时间和空间复杂度都比较低的拓扑排序算法。     

Efficient local search far DAG scheduling

Scheduling DAGs to multiprocessors is one of the key issues in high-performance computing. Most realistic scheduling algorithms are heuristic and heuristic algorithms often have room for improvement. The quality of a scheduling algorithm can be effectively improved by a local search. In this paper, we present a fast local search algorithm based on topological ordering. This is a compaction algorithm that can effectively reduce the schedule length produced by any DAG Scheduling algorithm. Thus, it can improve the quality of existing DAG scheduling algorithms. This algorithm can quickly determine the optimal search direction. Thus, it is of low complexity and extremely fast     

XForms in X-Smiles

World Wide Web Consortium is currently specifying XForms form language, which is intended to be the next generation language for WWW forms. In XForms, content is separated from presentation. Thus, forms can be versioned easily for different purposes. In addition, the form fields can be pre- and auto-filled. Finally, XForms allows the validation of the data at the client side before it is send to the server. The major browsers do not currently support XForms, which prevents the early utilisation of XForms. In this paper, we describe our implementation of XForms. The implementation is part of our X-Smiles browser, which is an open source XML browser.     

Disseminating dependent data in wireless broadcast environments

In wireless mobile environments, data broadcasting is an effective approach to disseminate information to mobile clients. In some applications, the access pattern of all the data can be represented by a weighted DAG. In this paper, we explore how to efficiently generate the broadcast schedule in a wireless environment for the data set having a weighted DAG access pattern. Such a broadcast schedule not only minimizes the access latency but also is a topological ordering of the DAG. Minimized access latency ensures the quality of service (QoS). We prove that it is NP-hard to find an optimal broadcast schedule and provide some heuristics. After giving an analysis for these heuristics on the latency and complexity, we implement all the proposed heuristics to compare their performance. Recommended by: Sunil Prabhakar     

Scheduling large-scale scientific workflow on virtual machines with different numbers of vCPUs

With the wide deployment of cloud computing in scientific computing, cost minimization is increasingly critical for large-scale scientific workflow. Unfortunately, due to the highly intricate directed acyclic graph (DAG)-based workflow and the flexible usage of virtual machines (VMs) in cloud platform, the existing workflow scheduling approaches are inefficient to strike a balance between the parallelism and the topology of the DAG-based workflow while using the VMs, which causes a low utilization of VMs and consumes more cost. To address these issues, this paper presents a novel task scheduling framework named cost minimization approach with the DAG splitting method (COMSE) for minimizing the cost of running a deadline-constrained large-scale scientific workflow. First, we provide comprehensive theoretical analyses on how to improve the utilization of a resource-balanced multi-vCPU VM for running multiple tasks simultaneously. Second, considering the balance between the parallelism and the topology of a workflow, we simplify the DAG-based workflow, and based on the simplified DAG, a DAG splitting method is devised to preprocess the workflow. Third, since the cloud is charged by hours, we also design an exact algorithm to find the optimal operation pattern for a given schedule to make the consumed instance hours minimum, and this algorithm is named as instance hours minimization by Dijkstra (TOID). Finally, by employing the DAG splitting method and the TOID, the COMSE schedules a deadline-constrained large-scale scientific workflow on the multi-vCPU VMs and incorporates two important objects: minimizing the computation cost and the communication cost. Our solution approach is evaluated through rigorous performance evaluation study using real-word workflows, and the results show that the proposed COMSE approach outperforms existing algorithms in terms of computation cost and communication cost.

     

基于拓扑序列归约的Web服务组合QoS度量算法

考虑有向无环图 (DAG)描述的组合服务模型,提出了一种新的组合服务QoS度量方法--基于拓扑序列归约的Web服务QoS度量方法（QCMTSR）。其借鉴迭代归约度量方法中的基本结构及QoS计算公式,定义了DAG图中的两类基本结构,串归约结构和并归约结构,并给出了两种基本结构的QoS属性计算公式;通过逐步归约DAG图拓扑序列中的每个节点,直至最后一个节点的QoS属性值就是组合服务的各QoS属性的度量结果。从理论上证明了QCMTSR算法适用于所有DAG描述的组合服务,并实验证明QCMTSR算法对可靠性和可用性能够更准确的度量。     

Teleform——种基于XML的Web界面描述模型

当前的form技术已经过时，满足不了一些特定应用的需要，文章介绍了W3C正在制定的基于XML的新一代form规范——XForms，并在XForms基础上提出了一种新的界面描述模型——Teleform，文章对XForms和Teleform进行了比较，最后提出了Teleform存在的一些不足以及未来的工作。     

基于最小化传输和完成时间的多DAG调度

针对云计算环境下多个有向无环图(DAG)工作流的调度问题,提出一种基于最小化数据传输时间和任务完成时间(LTCT)的算法,用于处理具有相同优先级的多个DAG工作流之间的调度问题。在多个DAG优先级各不相同时的情况下,给出多优先级多DAG的混合调度算法。实验结果表明,LTCT算法较E-Fairness算法在保证多DAG调度公平性的基础上,能避免额外的数据传输开销,有利于缩短整个工作流的执行Makespan,提高资源的利用率。