Abnormal user identification based on XGBoost algorithm

The eXtreme gradient boosting(XGBoost)algorithm is used to identify abnormal users.Firstly,the raw data were cleaned.Then user power characteristics were extracted from different aspects.Finally,the XGBoost classifier was used to identify the abnormal users respectively in the balanced sample set and the unbalanced sample set.In contrast,under the same characteristics,the k-nearest neighbor(KNN)classifier,back-propagation(BP)neural network classifier and random forest classifier were used to identify the abnormal users in the two samples.The experimental results show that the XGBoost classifier has higher recognition rate and faster running speed.Especially in the imbalanced data sets,the performance improvement is obvious.

Identification and Analysis of Multi-tasking Product Information Search Sessions with Query Logs

Purpose： This research aims to identify product search tasks in online shopplng ana analyze the characteristics of consumer multi-tasking search sessions. Design/methodology/approach： The experimental dataset contains 8,949 queries of 582 users from 3,483 search sessions. A sequential comparison of the Jaccard similarity coefficient between two adjacent search queries and hierarchical clustering of queries is used to identify search tasks. Findings： （1） Users issued a similar number of queries （1.43 to 1.47） with similar lengths （7.3-7.6 characters） per task in mono-tasking and multi-tasking sessions, and （2） Users spent more time on average in sessions with more tasks, but spent less time for each task when the number of tasks increased in a session. Research limitations： The task identification method that relies only on query terms does not completely reflect the complex nature of consumer shopping behavior.Practical implications： These results provide an exploratory understanding of the relationships among multiple shopping tasks, and can be useful for product recommendation and shopping task prediction. Originality/value： The originality of this research is its use of query clustering with online shopping task identification and analysis, and the analysis of product search session characteristics.

Potential of plant identification apps in urban forestry studies in China:comparison of recognition accuracy and user experience of five apps

Information on species composition of an urban forest is essential for its management.However,to obtain this information becomes increasingly difficult due to limited taxonomic expertise.In this study,we tested the possibility of using plant identification applications running on mobile platforms to fill this vacuum.Five plant identification apps were compared for their potential in identifying urban tree species in China.An online survey was conducted to determine the features of apps that contributed to users’satisfaction.The results show that identification accuracy varied significantly among the apps.The best performer achieved an accuracy of 74.6%at the species level,which is comparable to the accuracy by professionals in field surveys.Among the features of apps,accuracy of identification was the most important factor that contributed to users’satisfaction.However,plant identification apps did not perform well when used on rare species or outside of the regions where they have been developed.Results indicate that plant identification apps have great potential in urban forest studies and management,but users need to be cautious when deciding which one to use.

Multi Sensor-Based Implicit User Identification

Smartphones have ubiquitously integrated into our home and work environments,however,users normally rely on explicit but inefficient identification processes in a controlled environment.Therefore,when a device is stolen,a thief can have access to the owner’s personal information and services against the stored passwords.As a result of this potential scenario,this work proposes an automatic legitimate user identification system based on gait biometrics extracted from user walking patterns captured by smartphone sensors.A set of preprocessing schemes are applied to calibrate noisy and invalid samples and augment the gait-induced time and frequency domain features,then further optimized using a non-linear unsupervised feature selection method.The selected features create an underlying gait biometric representation able to discriminate among individuals and identify them uniquely.Different classifiers are adopted to achieve accurate legitimate user identification.Extensive experiments on a group of 16 individuals in an indoor environment show the effectiveness of the proposed solution:with 5 to 70 samples per window,KNN and bagging classifiers achieve 87–99%accuracy,82–98%for ELM,and 81–94%for SVM.The proposed pipeline achieves a 100%true positive and 0%false-negative rate for almost all classifiers.

Predicting and Classifying User Identification Code System Based on Support Vector Machines

In digital fingerprinting, preventing piracy of images by colluders is an important and tedious issue. Each image will be embedded with a unique User IDentification （UID） code that is the fingerprint for tracking the authorized user. The proposed hiding scheme makes use of a random number generator to scramble two copies of a UID, which will then be hidden in the randomly selected medium frequency coefficients of the host image. The linear support vector machine （SVM） will be used to train classifications by calculating the normalized correlation （NC） for the 2class UID codes. The trained classifications will be the models used for identifying unreadable UID codes. Experimental results showed that the success of predicting the unreadable UID codes can be increased by applying SVM. The proposed scheme can be used to provide protections to intellectual property rights of digital images aad to keep track of users to prevent collaborative piracies.

Automated Identification Algorithm Using CNN for Computer Vision in Smart Refrigerators

Machine Learning has evolved with a variety of algorithms to enable state-of-the-art computer vision applications.In particular the need for automating the process of real-time food item identification,there is a huge surge of research so as to make smarter refrigerators.According to a survey by the Food and Agriculture Organization of the United Nations(FAO),it has been found that 1.3 billion tons of food is wasted by consumers around the world due to either food spoilage or expiry and a large amount of food is wasted from homes and restaurants itself.Smart refrigerators have been very successful in playing a pivotal role in mitigating this problem of food wastage.But a major issue is the high cost of available smart refrigerators and the lack of accurate design algorithms which can help achieve computer vision in any ordinary refrigerator.To address these issues,this work proposes an automated identification algorithm for computer vision in smart refrigerators using InceptionV3 and MobileNet Convolutional Neural Network(CNN)architectures.The designed module and algorithm have been elaborated in detail and are considerably evaluated for its accuracy using test images on standard fruits and vegetable datasets.A total of eight test cases are considered with accuracy and training time as the performance metric.In the end,real-time testing results are also presented which validates the system’s performance.

User Session-Based Test Case Generation and Optimization Using Genetic Algorithm

An approach to generating and optimizing test cases is proposed for Web application testing based on user sessions using genetic algorithm. A large volume of meaningful user sessions are obtained after purging their irrelevant information by analyzing user logs on the Web server. Most of the redundant user sessions are also removed by the reduction process. For test reuse and test concurrency, it divides the user sessions obtained into different groups, each of which is called a test suite, and then prioritizes the test suites and the test cases of each test suite. So, the initial test suites and test cases, and their initial executing sequences are achieved. However, the test scheme generated by the elementary prioritization is not much approximate to the best one. Therefore, genetic algorithm is employed to optimize the results of grouping and prioritization. Meanwhile, an approach to generating new test cases is presented using crossover. The new test cases can detect faults caused by the use of possible conflicting data shared by different users.

基于Session的Web应用软件EFSM模型构建方法研究

在Web应用软件模型表示研究领域中,研究对象主要为不含Ajax技术的应用程序。少数针对Ajax(Asynchronous JavaScript and XML)的模型构建采用传统的FSM模型表示方法,并不能描述客户端消息触发后的参数传递问题;在FSM模型的基础上引入UML分层模型的表示方法因需要人工干预,不利于测试用例的自动生成。针对上述问题,借鉴EFSM模型这一重要的软件描述模型,从用户的Session数据出发,通过日志数据分析用户的行为,并记录客户端的操作事件。通过对用户行为和客户端操作事件进行匹配,生成完整的用户会话,从而建立EFSM模型。实验结果表明,该EFSM模型能够有效地表示Web应用程序的状态以及状态的变化情况,并可以有效地为测试用例自动生成服务。

用户偏好-制造商偏好双重视阈下的产品创新机会识别路径研究

精准识别产品创新机会对制造商规避盲目创新风险,持续获取竞争优势至关重要。针对现有产品创新机会识别研究中因视角单一致使制造商价值创造的效率难以最大化,以及识别的创新机会聚焦度不清晰等诸多问题,基于用户偏好制造商偏好双重视阈构建了产品创新机会识别路径。首先,运用结构主题模型从用户在线评论和产品描述文本中分别提取用户偏好及制造商偏好的产品属性主题;其次,基于两类主题之间的相似度进行主题分类以获取具象产品的创新需求;再次,从现有相关专利中提取创新要素并依托多维技术创新地图识别产品创新机会;最后,以老年智能手环靶向开展产品创新机会识别为例,验证了所提路径的可行性。应用结果表明,构建的用户偏好制造商偏好双重视阈下的产品创新机会识别路径能够为制造商高效开展产品创新活动提供可资借鉴的决策理论支持。

基于多特征符号聚合近似和层次聚类的户变关系识别方法

针对低压配电台区拓扑档案中可能存在的户变关系异常问题,文中提出了一种基于多特征符号聚合近似(MF-SAX)和层次聚类的户变关系识别方法。首先,运用符号聚合近似表达方法将用户电压时间序列转化为字符串序列,并引入电压波动系数和电压变化趋势两个附加参数对其特征表达进行强化。然后,基于编辑距离生成用户电压曲线相似性矩阵,并结合层次聚类算法实现户变关系的识别。最后,实际算例结果表明,提出的方法相比于现有方法准确率更高,误报更少,能直接应对数据缺失的情况,且具有更高的效率。

利用邻近度与内容特征的用户识别方法

社交网络对用户拓扑结构进行了获取限制,使得利用结构特征进行识别的方法准确率大大下降。利用邻近度与内容特征的用户识别方法构建了一个融合属性特征、结构特征与内容特征的基于XGboost的半监督网络模型,将跨社交网络用户识别问题转换为二分类问题。针对无法获得完整用户拓扑结构与种子用户不足的问题,提出显式好友与隐式好友的提取方法,根据待匹配用户对好友网络中的显式匹配用户对、隐式匹配用户对与其他好友将好友网络融合,结合用户重要度改进LINE算法二阶邻近度的经验概率,获取待匹配用户对的结构特征;将用户发文时间序列特征、生成内容关键词重叠度特征与关注用户标签特征作为生成内容特征;最后将属性特征、结构特征与内容特征进行融合完成用户识别。在真实数据集上的实验证明了本方法的有效性。

基于异质图注意力网络与多特征融合的跨社交媒体用户识别研究

跨社交媒体用户识别对于网络舆情的协同治理以及用户偏好的全方位识别与预测具有重要的指导意义。针对现有方法存在数据表达能力弱、忽略用户信息的动态性和关联性的问题,本文提出一种融合异质特征嵌入与实体动态关联的跨社交媒体用户识别模型。首先,整合用户的基本属性、生成内容和社交结构信息,构建各个社交媒体的异质信息网络;其次,通过设计新的元路径识别策略构造邻接矩阵,使用异质图注意力网络模型汇聚用户节点信息,增强节点特征的表示能力;再其次,引入了3种连续时间衰减函数,对跨社交媒体的实体相似矩阵进行加权,增强实体之间的动态关联;最后,融合单社交网络和跨社交网络中的以上特征,利用多层感知机实现跨社交媒体用户的识别和预测。在微博-知乎真实数据集中的研究结果显示,本文模型的整体性能优于其他基准模型,特别是线性衰减函数,其展现了最佳效果,且本文提出的元路径识别策略对提升用户识别效果具有重要作用。

副车架衬套分数阶粘弹动力学模型构建与应用研究

通过某车后副车架衬套动刚度试验数据分别辨识了分数阶Kelvin-Voigt模型和高阶分数阶导数FVMP模型,通过对比辨识拟合误差最终选择高阶分数导数三单元并联的衬套数学模型。基于Fortran语言和Adams中的用户子程序Gfosub文件,完成了橡胶衬套力学模型编写和编译,搭建了整车平顺性仿真平台,对提出的建模方法进行随机输入仿真试验。结果表明:线性模型的后座椅导轨加权加速度均方根值与试验值之间的误差大约为14.1%,新建模型的误差大约为9.2%,更接近实车平顺性评价指标,证明了模型的有效性。

透过“名”与“实”分离表象识别真实竞争情报行为场景--以香橼公司做空恒大地产报告为例

以香橼公司2012年做空恒大地产事件为例,提出透过“名”与“实”分离表象识别真实的竞争情报行为场景的学术问题。从竞争情报专业视角,概要介绍香橼公司2012年做空恒大地产报告内容要点,对香橼公司做空恒大地产案例进行解析,提出了对竞争情报研究与服务创新发展的启示,强调指出新形势下透过“名”与“实”分离表象识别现实社会中诸多真实的竞争情报行为场景,对于做好竞争情报研究与服务创新工作具有重要意义。

交通弱势出行者轨迹预测及安全管控策略

针对骑车人在交通事故中易受伤害的问题,提出一种基于骑车人意图识别的轨迹预测方法.首先,从骑车人角度及车辆角度提取骑车人的多元特征,如运动方向、回头概率、与车辆相对位置等,在十字路口场景下构建动态贝叶斯模型网络,分析了骑车人意图的影响因素,最终得到骑车人意图后验概率.其次,根据意图识别结果,构建了相关交通场景并给出了骑车人运动方程,通过基于粒子滤波方法的骑车人轨迹预测算法,结合运动方程及观测方程,以粒子模拟骑车人未来轨迹集.最后,搭建激光雷达和单目相机为主的数据采集硬件平台,获取与骑车人运动及姿态有关的信息,根据收集的7 358帧数据,计算出骑车人进入十字路口区域的时长,用以对算法进行评价.结果表明:本算法在骑车人过街前0.24~0.54 s可基本识别骑车人的意图,能够预测骑车人未来5.0 s内的轨迹集;算法具有较好的实用性,可以降低骑车人与车辆碰撞事故的发生概率.

基于AKNN异常检验与ADPC聚类的低压台区拓扑识别方法

低压台区拓扑信息的准确记录是进行台区线损分析、三相不平衡治理等工作的基础。针对目前拓扑档案排查成本高且效率低的问题,提出一种基于自适应k近邻(adaptive k nearest neighbor,AKNN)异常检验和自适应密度峰值(adaptive density peaks clustering,ADPC)聚类的低压台区拓扑识别方法。该方法利用动态时间弯曲(dynamic time warping,DTW)距离度量低压台区用户间电压序列的相似性,通过AKNN异常检验算法检验并校正异常的用户与变压器之间的关系(简称“户变关系”),在得到正确户变关系的基础上,采用ADPC聚类算法对台区内用户进行相位识别;最后,通过实际台区算例分析验证了该方法不需要人为设置参数,能有效实现低压台区的拓扑识别,具有较高的适用性与准确性。

基于多模态信息融合的用户连接方法

为弥补用户连接现有工作中存在的以下两个问题:大部分方法只考虑用户的文本和网络结构信息,忽略了图像等多模态信息;现有方法大都通过简单的特征拼接融合不同模态的信息,忽略了这些信息的互补性、关联性和异质性,提出一个基于多模态信息融合的用户连接模型MIFUIL。获取用户的文本、视觉以及异构网络的嵌入表示;利用注意力机制学习不同模态信息间的互补性、关联性和异质性,获取用户多模态信息的融合嵌入表示,通过多模态对比学习实现跨平台的用户连接。实验结果表明,MIFUIL模型在两个多模态数据集TWFQ、DB-YAGO上的性能均优于现有方法。

无线网络智能化的探索和应用

随着5G网络的完善和5G流量的快速增长,如何满足网络连接的多样化需求,提升客户感知保障能力,成为网络运营智能化的关键点。介绍了某省联通在无线网络智能化、潜在贬损用户识别与感知提升方面所做的探索与实践,通过整合XDR话单和专业网管等多源数据,引入熵权法、分层二元评分法等科学算法模型,结合用户体验回访结果不断迭代优化算法,实现了对潜在贬损用户的有效识别,同时在感知驱动网络问题闭环解决方面也取得了较好的效果。

融合多用户属性的网络知识社区核心用户识别研究:基于情感加权的LeaderRank算法

基于对网络知识社区用户属性的分析,本文提出一种融合多用户属性的核心用户识别方法,以提升核心用户识别的效率和效果,为提高社区运营和管理水平提供理论和方法参考。首先,基于用户的基本属性数据对用户的活跃性和专业性进行量化;其次,构建网络知识社区超网络模型,提出基于邻居好友重叠度的用户社交关系重要性算法、用户交互活动中的累计交互情感计算方法以及用户综合情感倾向性排名算法;最后,采用熵权法融合上述指标作为用户核心性得分,并通过得分排序识别核心用户。研究结果表明,相比于用户社交关系网络中的度中心性排名和用户交互关系网络中的LeaderRank排名,本文提出的融合多属性的网络知识社区核心用户识别方法具有更好的识别效果。

基于标幺化三阈值事件检测与LDA分类器的工商业负荷辨识方法

非侵入式负荷辨识技术能够低成本的获取用户各类设备使用情况,实现电力负荷的在线监测与分析,对支撑负荷预测、需求响应等应用开展有着重要意义。针对一般工商业用户类型多样、负荷种类繁多、设备运行特性复杂的特点,提出了一种基于标幺化三阈值事件检测与LDA分类器的工商业负荷辨识方案。首先针对不同能耗级别、不同启停特性的设备设计了参数可调的统一负荷事件检测框架,提升了缓变型、分段型、震荡型负荷事件的检出准确度。随后提出了基于多元特征与LDA线性判别的设备类型判断算法,在兼顾边缘端计算效率的同时取得了与随机森林等非线性分类器相同的辨识性能。