图池化作为图神经网络中重要的组件,在获取图的多粒度信息的过程中扮演了重要角色。而当前的图池化操作均以平等地位看待数据点,普遍未考虑利用邻域内数据之间的偏序关系,从而造成图结构信息破坏。针对此问题,本文提出一种基于偏序关系...图池化作为图神经网络中重要的组件,在获取图的多粒度信息的过程中扮演了重要角色。而当前的图池化操作均以平等地位看待数据点,普遍未考虑利用邻域内数据之间的偏序关系,从而造成图结构信息破坏。针对此问题,本文提出一种基于偏序关系的多视图多粒度图表示学习框架(multi-view and multi-granularity graph representation learning based on partial order relationships,MVMGr-PO),它通过从节点特征视图、图结构视图以及全局视图对节点进行综合评分,进而基于节点之间的偏序关系进行下采样操作。相比于其他图表示学习方法,MVMGr-PO可以有效地提取多粒度图结构信息,从而可以更全面地表征图的内在结构和属性。此外,MVMGr-PO可以集成多种图神经网络架构,包括GCN(graph convolutional network)、GAT(graph attention network)以及GraphSAGE(graph sample and aggregate)等。通过在6个数据集上进行实验评估,与现有基线模型相比,MVMGr-PO在分类准确率上有明显提升。展开更多
In order to balancing based on data achieve dynamic load flow level, in this paper, we apply SDN technology to the cloud data center, and propose a dynamic load balancing method of cloud center based on SDN. The appro...In order to balancing based on data achieve dynamic load flow level, in this paper, we apply SDN technology to the cloud data center, and propose a dynamic load balancing method of cloud center based on SDN. The approach of using the SDN technology in the current task scheduling flexibility, accomplish real-time monitoring of the service node flow and load condition by the OpenFlow protocol. When the load of system is imbalanced, the controller can allocate globally network resources. What＇s more, by using dynamic correction, the load of the system is not obvious tilt in the long run. The results of simulation show that this approach can realize and ensure that the load will not tilt over a long period of time, and improve the system throughput.展开更多
Many Task Computing(MTC)is a new class of computing paradigm in which the aggregate number of tasks,quantity of computing,and volumes of data may be extremely large.With the advent of Cloud computing and big data era,...Many Task Computing(MTC)is a new class of computing paradigm in which the aggregate number of tasks,quantity of computing,and volumes of data may be extremely large.With the advent of Cloud computing and big data era,scheduling and executing large-scale computing tasks efficiently and allocating resources to tasks reasonably are becoming a quite challenging problem.To improve both task execution and resource utilization efficiency,we present a task scheduling algorithm with resource attribute selection,which can select the optimal node to execute a task according to its resource requirements and the fitness between the resource node and the task.Experiment results show that there is significant improvement in execution throughput and resource utilization compared with the other three algorithms and four scheduling frameworks.In the scheduling algorithm comparison,the throughput is 77%higher than Min-Min algorithm and the resource utilization can reach 91%.In the scheduling framework comparison,the throughput(with work-stealing)is at least 30%higher than the other frameworks and the resource utilization reaches 94%.The scheduling algorithm can make a good model for practical MTC applications.展开更多
文摘图池化作为图神经网络中重要的组件,在获取图的多粒度信息的过程中扮演了重要角色。而当前的图池化操作均以平等地位看待数据点,普遍未考虑利用邻域内数据之间的偏序关系,从而造成图结构信息破坏。针对此问题,本文提出一种基于偏序关系的多视图多粒度图表示学习框架(multi-view and multi-granularity graph representation learning based on partial order relationships,MVMGr-PO),它通过从节点特征视图、图结构视图以及全局视图对节点进行综合评分,进而基于节点之间的偏序关系进行下采样操作。相比于其他图表示学习方法,MVMGr-PO可以有效地提取多粒度图结构信息,从而可以更全面地表征图的内在结构和属性。此外,MVMGr-PO可以集成多种图神经网络架构,包括GCN(graph convolutional network)、GAT(graph attention network)以及GraphSAGE(graph sample and aggregate)等。通过在6个数据集上进行实验评估,与现有基线模型相比,MVMGr-PO在分类准确率上有明显提升。
基金supported by the National Natural Science Foundation of China(No.61163058No.61201250 and No.61363006)Guangxi Key Laboratory of Trusted Software(No.KX201306)
文摘In order to balancing based on data achieve dynamic load flow level, in this paper, we apply SDN technology to the cloud data center, and propose a dynamic load balancing method of cloud center based on SDN. The approach of using the SDN technology in the current task scheduling flexibility, accomplish real-time monitoring of the service node flow and load condition by the OpenFlow protocol. When the load of system is imbalanced, the controller can allocate globally network resources. What＇s more, by using dynamic correction, the load of the system is not obvious tilt in the long run. The results of simulation show that this approach can realize and ensure that the load will not tilt over a long period of time, and improve the system throughput.
基金ACKNOWLEDGEMENTS The authors would like to thank the reviewers for their detailed reviews and constructive comments, which have helped improve the quality of this paper. The research has been partly supported by National Natural Science Foundation of China No. 61272528 and No. 61034005, and the Central University Fund （ID-ZYGX2013J073）.
文摘Many Task Computing(MTC)is a new class of computing paradigm in which the aggregate number of tasks,quantity of computing,and volumes of data may be extremely large.With the advent of Cloud computing and big data era,scheduling and executing large-scale computing tasks efficiently and allocating resources to tasks reasonably are becoming a quite challenging problem.To improve both task execution and resource utilization efficiency,we present a task scheduling algorithm with resource attribute selection,which can select the optimal node to execute a task according to its resource requirements and the fitness between the resource node and the task.Experiment results show that there is significant improvement in execution throughput and resource utilization compared with the other three algorithms and four scheduling frameworks.In the scheduling algorithm comparison,the throughput is 77%higher than Min-Min algorithm and the resource utilization can reach 91%.In the scheduling framework comparison,the throughput(with work-stealing)is at least 30%higher than the other frameworks and the resource utilization reaches 94%.The scheduling algorithm can make a good model for practical MTC applications.
