基于主题的用户兴趣域算法

针对用户兴趣偏好多变问题,提出一种兴趣特征权重随时间而变化的迭代计算方法。构造了用户兴趣特征与主题类间的二部图关系,并在此基础上提出了一种基于主题的用户兴趣聚类算法(TBC),改变了聚类对象"非此即彼"的硬划分方式。该算法所形成的基于主题的用户兴趣域结构,不仅充分表达了用户的多域兴趣特征和域间主题的联系,还能适应用户兴趣变化。实验表明,TBC算法比传统的K-Means算法以及属于软划分方式的FCM聚类具有更好的用户兴趣划分效果,并且在个性化推荐服务中表现出更高的推荐质量和效率。     

一种数据流处理环境下的节点副本放置方法

物联网环境下的许多应用表现为传感数据的连续流式处理,且系统往往通过节点的副本技术保障可用性。但是,运行时副本的备份和放置存在内存和带宽等资源开销,产生处理的延迟。该文给出一种方法,根据运行时的资源消耗以贪心方式放置节点的副本,折中了系统的可用性和开销。实际系统的仿真实验表明,在相同的条件下,该方法相比传统的随机放置,能为系统提供更稳定的可用性。     

面向MANET环境的动态自适应副本放置算法研究

移动自组网环境给数据复制技术带来了新的挑战。本文提出了一种分布式动态自适应副本放置算法,算法以数据访问的通信开销作为目标函数,每个副本节点从其邻居节点收集读写请求的局部信息,动态地调整副本放置方案以适应访问请求和网络拓扑的变化。本文还提出了基于稳定邻居的副本放置算法以减小由于节点移动给副本放置带来的振荡。模拟实验表明,在移动自组网环境中,本文提出的算法有效地减少了数据访问的通信开销,提高了数据访问的效率。     

个性化服务中用户兴趣聚类算法研究

讨论了个性化服务中用户兴趣建模对聚类算法的要求,指出经典聚类算法应用于用户兴趣聚类时的不足。在基于图论的K近邻聚类算法的基础上进行改进,提出一种基于相似度的聚类算法。实验证明,与K近邻算法相比,该算法能够显著提高聚类质量,有效区分孤立点,适用于用户兴趣聚类。     

基于XDR大数据分析和AI技术的定轨道路用户感知识别技术

随着网络结构、业务形态、用户模型的不断变化以及大数据、云计算等技术的不断演进,网络运维成本压力逐年上升,依赖于专业测试软件进行道路测试的传统优化模式已经难以满足未来发展的需要,网络优化亟待智能转型.基于定轨道路用户的真实海量数据,通过聚类算法模型对上海全路段用户感知问题进行定位分析研究.研究结果表明,大数据分析和AI算...     

基于用户兴趣分类优化的聚类模型仿真

对用户兴趣进行聚类分析对研究消费心理有着重要的意义.提出一种考虑用户兴趣分类优化的聚类模型,采用ID3决策树算法提高用户兴趣分类计算速度,将最高信息增益的属性当成前节点的检测属性,确保结果分解中的用户兴趣样本分类所需的信息量最小,构建用户兴趣分类优化的自适应模糊聚类目标函数,更新聚类原型矩阵,在自适应模糊聚类模型下,直接给出聚类原型的迭代等式,保证分类准确.实验结果说明,所提模型相对于传统聚类模型不容易陷入局部最优解,具有较高的查全率和查准率,对进一步用户行为研究有着较大的意义.     

基于用户感知体系的神经网络算法研究

用户感知是目前无线移动网络研究的热点。主要讨论了人工神经网络算法运用在用户感知体系上的情况。通过用户感知体系发展特点及要求，对几种用户感知体系进行比较。最终将神经网络算法与改进后用户感知体系相结合，达到优化体系映射的目的。     

基于用户兴趣和双重聚类融合的协同过滤算法的优化研究

协同过滤技术是目前应用最多的一种推荐技术,这项技术从用户提供的信息中展开发掘,按"物以类聚,人以群分"的原则产生和目标用户(或项目)相似性高的最近邻,从中预测评分,进而产生推荐。但是由于评分信息稀疏化就会造成无法适应用户兴趣,而且推荐的实时性差等问题。针对上述问题,文章提出了一种带有改进的用户-项目类型喜好相似性的计算方法完善用户兴趣改变的问题,并且结合了优化后的双重k-means聚类,使搜索最近邻的范围大大减少,从而提高了推荐算法的实时性。实验结果表明,该优化后的协同过滤推荐算法能通过时间相似性更好地适应用户兴趣的变化,推荐的精度最精确,效果更易使用户满意。     

基于聚类的隐式用户兴趣建模

随着Internet的迅速发展,人们必须面对信息爆炸的现实。描述了一种关键词向量的方式表达用户兴趣。将BIRCH聚类算法应用于用户访问的网络文档上来建立用户兴趣模型。基于Myspace用户日志,又实现了一个用户兴趣建模系统,该系统验证了提出方法的有效性。     

基于近邻传播算法的K-means聚类优化算法

K-means聚类算法在随机选择的初始聚类中心的基础上进行聚类,其聚类效果会因为初始聚类中心的不确定性而不稳定。为了优化其聚类效果,提出了基于近邻传播算法(AP算法)的K-means聚类优化算法(APK-means)。该算法首先通过近邻传播算法生成若干个初始聚类,然后依序选择k个聚类规模最大的聚类中心作为K-means聚类算法的初始聚类中心,接着运行K-means聚类。算法有效性分析和实验结果验证了该算法有效优化了K-mean算法的聚类稳定性和有效性。     

QoS-aware replica placement for data intensive applications

In recent years, with the rapid development of data intensive applications, data replication has become an enabling technology for the data grid to improve data availability, and reduce file transfer time and bandwidth consumption. The placement of replicas has been proven to be the most difficult problem that must be solved to realize the process of data replication. This paper addresses the quality of service （QoS） aware replica placement problem in data grid, and proposes a dynamic programming based replica placement algorithm that not only has a QoS requirement guarantee, but also can minimize the overall replication cost, including storage cost and communication cost. By simulation, experiments show that the replica placement algorithm outperforms an existing popular replica placement technique in data grid.     

Using optorsim to efficiently simulate replica placement strategies

Optorsim provides simulations of file replication strategies such as replica placement,replication scheduling,replica consistency maintenance,etc.However,to assess a replica strategy a researcher must write network configuration files and repeatedly modify parameters,which is inefficient.In this article,a scale-free algorithm is developed to generate network topology which is loaded into Optorsim,and a graphical user interface(GUI) is implemented to set all configuration parameters and algorithm parameters....     

Distance control strategy for deploying content replica servers in an edge cloud

To reduce fetching cost from a remote source,it is natural to cache information near the users who may access the information later.However,with development of 5 G ultra-dense cellular networks andmobile edge computing(MEC),a reasonable selection among edge servers for content delivery becomes a problem when the mobile edge obtaining sufficient replica servers.In order to minimize the total cost accounting for both caching and fetching process,we study the optimal resource allocation for the content replica servers’ deployment.We decompose the total cost as the superposition of cost in several coverages.Particularly,we consider the criterion for determining the coverage of a replica server and formulate the coverage as a tradeoff between caching cost and fetching cost.According to the criterion,a coverage isolation(CI) algorithm is proposed to solve the deployment problem.The numerical results show that the proposed CI algorithm can reduce the cost and obtain a higher tolerance for different centrality indices.     

Hybrid genetic-paired-permutation algorithm for improved VLSI placement

This paper addresses Very large-scale integration (VLSI) placement optimization, which is important because of the rapid development of VLSI design technologies. The goal of this study is to develop a hybrid algorithm for VLSI placement. The proposed algorithm includes a sequential combination of a genetic algorithm and an evolutionary algorithm. It is commonly known that local search algorithms, such as random forest, hill climbing, and variable neighborhoods, can be effectively applied to NP-hard problem-solving. They provide improved solutions, which are obtained after a global search. The scientific novelty of this research is based on the development of systems, principles, and methods for creating a hybrid (combined) placement algorithm. The principal difference in the proposed algorithm is that it obtains a set of alternative solutions in parallel and then selects the best one. Nonstandard genetic operators, based on problem knowledge, are used in the proposed algorithm. An investigational study shows an objective-function improvement of 13%. The time complexity of the hybrid placement algorithm is O(N²).     

Modeling chunk-based content placement in information centric networking

Content-centric networking (CCN) proposes a content-centric paradigm which changes the waist hourglass from Internet protocol (IP) to content chunk. In this paper, based on content chunks, an optimization model of minimizing the total delay time in information centric networking (ICN) is established, and branch-and-bound method and greedy (BG) algorithm is proposed to get the content placement method. As the multipath is natural supported in CCN, chunk-based content placement can decline delay time obviously, even it would increase the calculation amount which can be solved easily by the node's capacity. Simulation results indicate that the chunk-based content placement scheme is better than the single-based cache policy on the network total delay time, and the best number of each content chunk split is decided by the link density and the number of the nodes in the network.     

SaaS多租户数据放置的优化调整策略研究

采用副本技术的SaaS多租户数据为满足租户访问及服务提供商管理的需求,在云中多节点必须合理放置。针对多租户数据特征和节点负载情况,对负载超重节点或超轻节点,通过调整数据副本数目和位置进行放置策略维护及优化,在满足所有租户SLA需求的同时最小化服务提供商的成本代价,实现服务质量与管理成本2个目标的均衡。并通过实验与随机放置策略和贪婪放置策略进行比较,证明该策略的可行性和有效性。     

An efficient macro-cell placement algorithm

A new approximation algorithm is presented for the efficient handling of large macro-cell placement problems. The algorithm combines simulated annealing with new features based on a hierarchical approach and a divide-and-conquer technique. Numerical results show that these features can lead to a considerable increase in efficiency of the placement algorithm without loss of effectiveness.     

基于概率的启发式ICN缓存内容放置方法

提出一种基于概率存储的启发式住处中心网络内容缓存方法（PCP）。主要思想是请求消息和数据消息在传输过程中统计必要信息,当数据消息返回时,沿途各缓存节点按照一定概率决策是否在本地缓存该内容。设计缓存概率时综合考虑内容热度和缓存放置收益,即内容热度越高,放置收益越大的内容被缓存的概率越高。实验结果表明,PCP在缓存服务率、缓存命中率、平均访问延迟率等方面,与现有方法相比具有显著优势,同时PCP开销较小。     

基于优化粒计算下微粒子动态搜索的K-medoids聚类算法

K?medoids算法具有对初始聚类中心敏感,聚类准确度不高及时间复杂度大的缺点。基于此,文中提出一种优化的K?medoids算法;该算法在已有的粒计算初始化基础上进行了改进,以对象之间的相似性作为判断依据,结合最大最小法初始化聚类中心,能有效地获取最佳或近似最佳的聚类中心;在优化的粒计算前提下,提出了基于微粒子动态搜索策略,以初始中心点作为基点,粒子内所有对象到其中心的平均距离为半径,形成一个微粒子;在微粒子内部,采用离中心点先近后远的原则进行搜索,能有效地缩小搜索范围,提高聚类准确率。实验结果表明：在UCI多个标准数据集中测试,且与其他改进的K?medoids算法比较分析,该算法在有效缩短收敛时间的同时保证了算法聚类准确率。     

网络延迟聚类的宏观预警的检测点放置

大规模异常事件爆发,使网络可用性受到严重威胁。引发人们对大规模入侵检测系统的研究,其检测效果直接取决于网络检测点选取。大规模网络预警检测点放置问题被转化为网络拓扑有权图聚类问题。针对层次聚类算法面临的初始点选择问题提出基于出度分离初始点选择算法,降低聚类结果对初值的依赖。同时,提出改进双向层次聚类算法DHC,有效降低了结果簇数量。实验证明,此算法整体最优并成功应用于测得的真实全国骨干网络,有效解决了检测点放置问题。