| 基于无限深度神经网络的非平衡大数据集群匿名化调度算法 |
 点此下载全文 |
| 引用本文：张 婷.基于无限深度神经网络的非平衡大数据集群匿名化调度算法[J].计算技术与自动化,2022,(3):71-76 |
| 摘要点击次数: 228 |
| 全文下载次数: 0 |
|
|
| 中文摘要:以提升非平衡大数据集群数据调度的高效性、准确性、安全性为目的，提出基于无限深度神经网络的非平衡大数据集群匿名化调度算法。数据采集模块利用传感器采集非平衡大数据集群中各节点服务器的数据信息，传输至数据分类模块的深度最小最大概率机中，通过卷积神经网络的端对端方式，训练获得的非平衡大数据，提取原始大数据高级特征，以最小最大概率机作为分类器，实现非平衡大数据的分类；应用数据匿名化模块的K-匿名化算法匿名化处理分类后的数据信息，经数据调度模块中基于蚁群算法的负载均衡调度方法，进行非平衡大数据集群各节点、任务的实时调度。实验结果表明：该算法可提升非平衡大数据的分类准确度，降低信息损失率、泄密风险，并可充分调度优势节点与空闲节点实现任务的均衡分配，提升非平衡大数据集群的任务调度能力。 |
| 中文关键词:无限深度神经网络 非平衡大数据 匿名化 卷积神经网络 最小最大概率机 负载均衡 |
| |
| Anonymous Scheduling Algorithm for Unbalanced Big Data Cluster Based on Infinite Depth Neural Network |
|
|
| Abstract:In order to improve the efficiency, accuracy and security of data scheduling of unbalanced big data cluster, an anonymous scheduling algorithm of unbalanced big data cluster based on infinite depth neural network is proposed. The data acquisition module uses the sensor to collect the data information of each node server in the unbalanced big data cluster and transmit it to the depth minimum maximum probability machine of the data classification module. Through the end-to-end method of convolution neural network, the unbalanced big data obtained is trained, the advanced features of the original big data are extracted, and the minimum maximum probability machine is used as the classifier, Realize the classification of unbalanced big data; The k-nonymization algorithm of the data anonymization module is applied to anonymously process the classified data information, and the load balancing scheduling method based on Ant Colony Algorithm in the data scheduling module is used to schedule the nodes and tasks of the unbalanced big data cluster in real time. The experimental results show that the algorithm can improve the classification accuracy of unbalanced big data, reduce the rate of information loss and the risk of disclosure, fully schedule advantageous nodes and idle nodes, realize the balanced allocation of tasks, and improve the task scheduling ability of unbalanced big data clusters. |
| keywords:infinite depth neural network unbalanced big data anonymization convolutional neural network minimum maximum probability machine load balancing |
| 查看全文 查看/发表评论 下载pdf阅读器 |
次