Maximizing Influence in Temporal Social Networks:A Node Feature-Aware Voting Algorithm

Influence Maximization(IM)aims to select a seed set of size k in a social network so that information can be spread most widely under a specific information propagation model through this set of nodes.However,most existing studies on the IM problem focus on static social network features,while neglecting the features of temporal social networks.To bridge this gap,we focus on node features reflected by their historical interaction behavior in temporal social networks,i.e.,interaction attributes and self-similarity,and incorporate them into the influence maximization algorithm and information propagation model.Firstly,we propose a node feature-aware voting algorithm,called ISVoteRank,for seed nodes selection.Specifically,before voting,the algorithm sets the initial voting ability of nodes in a personalized manner by combining their features.During the voting process,voting weights are set based on the interaction strength between nodes,allowing nodes to vote at different extents and subsequently weakening their voting ability accordingly.The process concludes by selecting the top k nodes with the highest voting scores as seeds,avoiding the inefficiency of iterative seed selection in traditional voting-based algorithms.Secondly,we extend the Independent Cascade(IC)model and propose the Dynamic Independent Cascade(DIC)model,which aims to capture the dynamic features in the information propagation process by combining node features.Finally,experiments demonstrate that the ISVoteRank algorithm has been improved in both effectiveness and efficiency compared to baseline methods,and the influence spread through the DIC model is improved compared to the IC model.

Convolutional Neural Network-Based Identity Recognition Using ECG atDifferent Water Temperatures During Bathing

This study proposes a convolutional neural network(CNN)-based identity recognition scheme using electrocardiogram(ECG)at different water temperatures(WTs)during bathing,aiming to explore the impact of ECG length on the recognition rate.ECG data was collected using non-contact electrodes at five different WTs during bathing.Ten young student subjects(seven men and three women)participated in data collection.Three ECG recordings were collected at each preset bathtub WT for each subject.Each recording is 18 min long,with a sampling rate of 200 Hz.In total,150 ECG recordings and 150 WT recordings were collected.The R peaks were detected based on the processed ECG(baseline wandering eliminated,50-Hz hum removed,ECG smoothing and ECG normalization)and the QRS complex waves were segmented.These segmented waves were then transformed into binary images,which served as the datasets.For each subject,the training,validation,and test data were taken from the first,second,and third ECG recordings,respectively.The number of training and validation images was 84297 and 83734,respectively.In the test stage,the preliminary classification results were obtained using the trained CNN model,and the finer classification results were determined using the majority vote method based on the preliminary results.The validation rate was 98.71%.The recognition rates were 95.00%and 98.00%when the number of test heartbeats was 7 and 17,respectively,for each subject.

Majority-Vote on Undirected Barab´asi-Albert Networks

On Barab´asi-Albert networks with z neighbours selected by each added site,the Ising model was seen to show a spontaneous magnetisation.This spontaneous magnetisation was found below a critical temperature which increases logarithmically with system size.On these networks the majority-vote model with noise is now studied through Monte Carlo simulations.However,in this model,the order-disorder phase transition of the order parameter is well defined in this system and this was not found to increase logarithmically with system size.We calculate the value of the critical noise parameter qc for several values of connectivity z of the undirected Barab´asiAlbert network.The critical exponentesβ/ν,γ/νand 1/νwere also calculated for several values of z.

基于互信息与萤火虫算法的网络入侵特征选择

为减少网络入侵检测数据中的冗余特征,提出一种结合互信息和萤火虫算法的特征选择方法。针对互信息不能精确计算特征间冗余度,提出类内特征冗余互信息特征选择方法。针对萤火虫算法步长因子固定易使算法陷入局部最优等问题,提出自适应步长萤火虫算法特征选择。以上方法分别选取特征子集后利用投票策略选取最优子集,对该子集基于C4.5和贝叶斯网络分类器分类。实验结果表明,使用10个特征检测能有效提高入侵检测率、误报率和F-measure,同时还缩短训练和检测时间。此外,与现有的几种方法相比,该方法在准确率、检测率和F-measure都获得不错效果。

具有层级结构集体影响力的多数投票模型

多数投票模型是观点动力学研究中的常用模型,本文在多数投票模型的基础上引入了具有层级结构的集体影响力,以节点周边层级结构上的节点的度衡量中心节点的观点权重,即为集体影响力参数.通过蒙特卡罗模拟,研究了具有集体影响力的多数投票模型在ER(Erdos and Rényi)随机网络与无标度网络上观点的演化,发现系统观点均出现了有序-无序相变,且相比原始多数投票模型更容易趋于无序,即相变临界点更小.原因是考虑具有层级结构的集体影响力时,系统的集体影响力参数值整体减小,且分布数目随着参数值的增大而减少,呈“长尾”趋势,占少数的高影响力个体使周围节点的观点产生跟随现象,随着噪声参数的增大,当少数的高影响力个体趋于无序时,整个系统也会趋于无序,即系统更容易达到无序状态.最后通过有限尺寸标度法,发现无论在ER随机网络或在无标度网络中,具有集体影响力的多数投票模型的相变均为Ising模型普适类.

Predicting mobile users’behaviors and locations using dynamic Bayesian networks

This paper studies the traveling location prediction problem for detecting whether mobile users will leave their living area and where they will go.We investigate the hidden connections between users’behaviors in different locations and online social interactions.We combine dynamic Bayesian networks with a majority voting model which is based on social interaction information to estimate the users’behaviors and predict the locations.By analyzing Instagram media records,spanning a period of 3 months,we explore rarely visited locations,which are often ignored as noise in previous research.In comparison,our model,using Instagram data with two existing location prediction models,shows that(1)our location prediction is more accurate and robust in both the general location and the location outside the living area;(2)social relations are instrumental in the location prediction as social interaction information can increase the accuracy of the prediction.

基于深度网络投票的抗血管生成肽识别

血管生成在各种疾病中,尤其是癌症的发病机制中起着关键作用,因此开发更加快速高效的抗血管生成肽(AAPs)智能识别工具尤为重要.基于多种特征工程、深度学习和集成学习构建了一个深度网络投票的识别模型iAAPs-DNV.采用AAindex编码、分组权重编码(EBGW)、K-间隔氨基酸对(KSAAP)、基于物理化学性质的二阶移动平均(SOMA)和BLOSUM62编码提取氨基酸序列的特征信息.利用软投票策略集成加入了注意力机制(attention)的双向长短期记忆网络(BiLSTM)和卷积神经网络(CNN),并通过全连接层输出识别结果.iAAPs-DNV模型在Main数据集和NT15数据集上的识别精度明显优于已有的识别模型,表明该模型能够高效准确地识别抗血管生成肽.

基于IRF-PF算法的WLAN室内定位方法

针对室内复杂环境下无线局域网接收信号强度的时变波动性,提出一种基于改进随机森林算法并融合粒子滤波的室内定位方法。首先,通过自适应滑动窗口选取接收信号强度较为平滑的信号接入点作为定位指纹特征建立决策树,从而减少特征空间的噪声;然后,利用bagging抽样后的袋外数据赋予决策树不同的投票权重等级,增大决策树类别间判别差异性,优化随机森林算法在目标位置与接收信号强度间弱映射关联性场景下定位精度低的问题,依据改进的随机森林算法建立位置坐标与接收信号强度间的映射模型,计算定位点的位置;最后,为了提高定位准确性与稳定性,将依据改进随机森林算法计算出的位置作为粒子滤波的观测量,利用粒子滤波对目标位置进行估计和预测。在Zenodo数据集上的实验结果表明:提出的算法在定位误差2 m范围内精度达到75.8%,在4 m范围内精度已经高达94.8%;相比于传统随机森林算法,能够将2 m内的定位精度提高11.5%;与K近邻、岭回归和贝叶斯等室内定位算法相比,定位精度和算法实时性也得到明显提升。

面向英文阅读难度分类的神经网络设计与实现

阅读文本是影响英语阅读材料难度的重要因素,有效地评估英文文本难度可以为不同阅读能力的学习者提供相匹配的阅读材料,帮助教师科学选择合适的教学资源,为考试命题者提供科学指导。基于神经动力学方法(NDA)的收敛差分神经网络(CDNN),将文本数据进行特征选择、加权和样本归一化预处理,用于训练不同映射函数的网络,再将网络的输出通过投票规则进行增强泛化,从而实现了一种结合投票收敛差分神经网络(Voting-CDNN,V-CDNN)的英文文本阅读难度分类方法,提高了计算效率和分类预测准确率。实验结果表明,V-CDNN的分类准确率最高值和平均值分别达到98.81%和95.45%,其在计算时间、平均精度和最高精度等方面进一步证实了V-CDNN是一个高性能的分类器。

结合特征选择的抗噪声网络流分类

在现代网络管理中,网络环境愈加复杂,网络流噪声已经成为不可忽视的因素,然而现有的抗噪声流分类方法在实际效果上仍有不尽人意之处。针对这一问题,本文提出了一种结合特征选择的抗噪声网络流分类方法NNTC-FS,该方法采用投票机制判定噪声并搭建级联结构,实现先过滤再分类的线上任务。在公共数据集上的实验表明,NNTC-FS能实现90%以上的分类正确率,并在分类精度和时间性能上优于文献方法。

Network-Based Voting System—Evaluation and Optimization of CCMVS

This paper gives a brief introduction to a novel voting system, the Network-based Voting System (NVS). The system design is based on the careful analysis and evaluation of a traditional voting system, the computer controlled and managed voting system. The new system integrates technologies such as image processing, networking and databases to enhance three aspects of system performance: data collection, data transfer, and data management. Experiments have proved that the performance of the network-based voting system is superior to the CCMVS.

Deep Learning-Based Robust Morphed Face Authentication Framework for Online Systems

The amalgamation of artificial intelligence(AI)with various areas has been in the picture for the past few years.AI has enhanced the functioning of several services,such as accomplishing better budgets,automating multiple tasks,and data-driven decision-making.Conducting hassle-free polling has been one of them.However,at the onset of the coronavirus in 2020,almost all worldly affairs occurred online,and many sectors switched to digital mode.This allows attackers to find security loopholes in digital systems and exploit them for their lucrative business.This paper proposes a three-layered deep learning(DL)-based authentication framework to develop a secure online polling system.It provides a novel way to overcome security breaches during the face identity(ID)recognition and verification process for online polling systems.This verification is done by training a pixel-2-pixel Pix2pix generative adversarial network(GAN)for face image reconstruction to remove facial objects present(if any).Furthermore,image-to-image matching is done by implementing the Siamese network and comparing the result of various metrics executed on feature embeddings to obtain the outcome,thus checking the electorate credentials.

基于集成深度强化学习的自动驾驶车辆行为决策模型

提出一种基于集成的深度强化学习的自动驾驶车辆的行为决策模型。基于Markov决策过程(MDP)理论,采用标准投票法,将深度Q学习网络(DQN)、双DQN(DDQN)和竞争双DDQN(Dueling DDQN)等3种基础网络模型集成。在高速公路仿真环境、在单向3车道、4车道、5车道数量场景下,对向左换道、车道保持、向右换道、同车道加速和减速等5种车辆驾驶行为,进行测试和泛化性验证。结果表明:与其它3种网络模型相比,该模型的决策成功率分别提高了6%、3%和6%;平均车速也有提升;100回合的测试,耗时小于1 ms,满足决策实时性要求。因而,该决策模型提高了行车安全和决策效率。

基于集成学习双流神经网络的实时面部篡改视频检测模型

恶意面部篡改对社会安全和稳定存在负面影响,对面部篡改后的视频图像进行准确的检测是一个十分重要的课题。为了解决视频检测模型实时性较差的问题,提出一种基于集成学习双流循环神经网络的面部篡改视频检测模型,并引入集成学习中的投票机制。首先,接收少量连续的序列帧,通过卷积神经网络进行空间特征的提取,同时引入中心差分卷积进行空间域的篡改伪影增强。然后,将连续的序列帧进行差分,以增强时间域上的篡改伪影,同时通过卷积神经网络进行时间特征的提取。随后,将空间域和时间域的双流特征向量进行拼接,通过循环神经网络进行特征提取。在循环神经网络特征提取过程中,逐帧的特征信息被保留下来作为后续辅助帧级分类器的输入,同时循环神经网络的最终输出作为视频级判别器的输入。最后,引入集成模型的投票机制整合多个辅助帧级判别器和视频级判别器的输出,并通过引入权重超参数γ来平衡辅助帧级判别器和视频级判别器的重要程度,帮助模型提高检测准确率。在FaceForensics++数据集上,与主流检测模型进行对比,所提模型平均准确率提升了0.4%和1.0%。同时,所提模型可以仅使用较少连续帧进行篡改检测,提高了模型的实时性。

反应堆保护系统网络通信设计的研究与优化

反应堆保护系统保障机组在紧急情况下实现反应堆紧急停堆及专设安全设施驱动控制连锁,是关系机组安全的重要仪控系统。通过对网络通信板卡在反应堆保护系统中的作用进行研究,发现在当前设计下存在的安全隐患。通过对保护逻辑及硬件设计的进一步研究,提出了相应的优化方案,以降低网络通信板卡故障造成专设安全设施误动的风险,达到提高反应堆保护系统运行稳定性的目的。

基于Voting机制的IMA-BP不平衡数据分类算法

针对传统分类模型在实际应用中对提取到的不平衡数据特征进行分类时分类结果精度低的问题,提出使用蜉蝣算法(mayfly algorithm,MA)优化的反向传播(back propogation,BP)神经网络分类模型。同时为了提升算法前期全局搜索能力和后期局部搜索能力,引入阻尼比系数和非线性惯性权重因子,构建出改进蜉蝣算法(improved mayfly algorithm,IMA)优化的BP神经网络(IMA-BP)分类器。根据该分类器分类具有随机的特点,引入集成学习中的投票(Voting)机制,将IMA-BP作为弱分类器,将各弱分类器的分类结果通过软投票方法融合,构成了一个Voting机制的IMA-BP分类模型。为验证分类模型的性能,使用UCI数据库中的数据集将该模型与其他的模型进行比较,结果表明Voting机制的IMA-BP分类模型对4个数据集的分类准确率分别为88.67%、96.67%、91.25%、93.52%,都要高于其他模型,说明该分类模型具有较好准确性和可行性,对一些分类任务具有较强的指导作用和应用价值。

基于机器学习的SDN异常流量检测架构

针对SDN网络中控制器容易受到DDoS攻击导致CPU资源耗尽的问题,提出一种基于机器学习的SDN异常流量检测架构。根据DDoS攻击在通信、频率等方面的特性从流表中提取相关联的七维特征,使用互信息法筛选出四维最优特征子集,结合集成投票算法检测异常流量,利用SDN转控分离的独特性质提出多目标流路由方案,为正常流量分配高带宽、低延迟的优化路径。实验结果表明,提出架构能及时准确检测到DDoS攻击,集成投票算法在时间开销和CPU平均利用率方面有较好改善。

考虑碳排放的综合能源系统抗差参数辨识

近年来随着综合能源系统研究的深入,对能源设备建模的完整性和准确性提出了更高要求。针对传统研究中尚未考虑设备碳排放参数以及现有参数辨识方法的不足,文章提出一种考虑碳排放的综合能源系统抗差参数辨识方法。首先,建立了考虑碳排放强度的常见能源设备模型;其次,针对获取的系统历史运行数据,利用BP神经网络对其中的缺失值进行拟合填充,获取高可信的有效数据;最后,采用基于投票法的参数辨识方法对设备进行参数辨识,并通过旧数据遗忘与更新对设备参数进行动态辨识,通过算例对比分析了所提方法相较于现有方法的优势,说明了所提方法的必要性和有效性。

基于多卷积神经网络融合的SAR舰船分类

针对SAR图像中小型舰船分类准确率较低的问题,提出一种多卷积神经网络加权融合的方法。首先构建高分辨率卷积神经网络对特征图进行多尺度融合,引入微调模型和标签平滑减少训练过拟合的问题;然后利用高分辨网络、MobileNetv2网络和SqueezeNet网络训练3种单分类模型;最后采用加权投票方式对3种分类模型的结果进行融合。采用融合算法对GF-3号舰船数据集进行分类实验,取得94.83%的准确率、95.43%的召回率和0.9513的F1分数的分类性能。实验结果表明,该舰船分类算法模型具有较优的分类能力,验证了其在高分辨率SAR图像舰船分类上的有效性。

基于小样本利用神经网络组合评价软件质量

分析用于软件质量评价的神经网络在训练时只有小样本的原因,及由此导致的网络泛化能力差和泛化能力无法定量估算等问题,提出输出一致假设,用“分析评价结果和表决结果”作为解决问题的途径,以已经训练出对小样本准确率高且未过拟合的少量相对合格网络为前提,依据蒙特卡洛模拟方法生成的大量随机数,采用表决成功率(VSR)和评价相同率均值(SR)反推出泛化能力最优的网络组合,用t分布定量估算最优网络组合泛化能力。实例结果表明,该方法基于小样本,能够挑选出对新软件项目泛化能力较强的最优网络组合,为从软件度量直接评价软件综合质量的工作和集成学习提供新思路。