检索结果-南通市图书馆

MMCo:using multimodal deep learning to detect malicious traffic with noisy labels

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2024年第1期18卷 239-241页

作者： Qingjun YUAN Gaopeng GOU yuefei zhu Yongjuan WANG Strategic Support Force Information Engineering University Zhengzhou 450001China Institute of Information Engineering Chinese Academy of SciencesBeijing 100093China

1 Introduction The success of a deep learning-based network intrusion detection systems(NIDS)relies on large-scale,labeled,realistic traffic[1,2].However,automated labeling of realistic traffic,such as by sandbox and rule-based approaches,is prone to errors[3],which in turn affects deep learning-based *** effective schemes for learning with noisy labels(LNL)have been proposed,among which the use of parallel networks for sample selection during training has been shown to be *** is argued that parallel networks can discriminate samples from different perspectives,thus adding additional information to the training *** centerpiece is maintaining disagreement of networks.

关键词： network networks modal

Combo Packet:An Encryption Traffic Classification Method Based on Contextual Information

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第4期79卷 1223-1243页

作者： Yuancong Chai yuefei zhu Wei Lin Ding Li State Key Laboratory of Mathematical Engineering and Advanced Computing Information Engineering UniversityZhengzhou450001China

With the increasing proportion of encrypted traffic in cyberspace, the classification of encrypted traffic has becomea core key technology in network supervision. In recent years, many different solutions have emerged in this *** methods identify and classify traffic by extracting spatiotemporal characteristics of data flows or byte-levelfeatures of packets. However, due to changes in data transmission mediums, such as fiber optics and satellites,temporal features can exhibit significant variations due to changes in communication links and transmissionquality. Additionally, partial spatial features can change due to reasons like data reordering and *** with these challenges, identifying encrypted traffic solely based on packet byte-level features is significantlydifficult. To address this, we propose a universal packet-level encrypted traffic identification method, ComboPacket. This method utilizes convolutional neural networks to extract deep features of the current packet andits contextual information and employs spatial and channel attention mechanisms to select and locate effectivefeatures. Experimental data shows that Combo Packet can effectively distinguish between encrypted traffic servicecategories (e.g., File Transfer Protocol, FTP, and Peer-to-Peer, P2P) and encrypted traffic application categories (e.g.,BitTorrent and Skype). Validated on the ISCX VPN-non VPN dataset, it achieves classification accuracies of 97.0%and 97.1% for service and application categories, respectively. It also provides shorter training times and higherrecognition speeds. The performance and recognition capabilities of Combo Packet are significantly superior tothe existing classification methods mentioned.

关键词： Encrypted traffic classification packet-level convolutional neural network attention mechanisms

一种基于AdaBoost模型的区块链异常交易检测方案

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

信息网络安全 2024年第1期 24-35页

作者：宋玉涵祝跃飞魏福山中国人民解放军信息工程大学网络空间安全学院郑州450001

针对区块链系统加密货币交易记录中存在的盗币异常行为,文章基于Ada Boost模型提出一种具有隐私保护功能的异常交易检测方案。该方案采用加法同态加密和矩阵混淆技术,在有效识别并预测异常交易的同时,保证交易数据的隐私性。此外,在云... 详细信息

针对区块链系统加密货币交易记录中存在的盗币异常行为,文章基于Ada Boost模型提出一种具有隐私保护功能的异常交易检测方案。该方案采用加法同态加密和矩阵混淆技术,在有效识别并预测异常交易的同时,保证交易数据的隐私性。此外,在云外包环境中设计实现方案的底层协议,并证明了方案的正确性和隐私保护性质。与同类协议相比,该方案在保证隐私性的同时,具有较高的检测准确率和召回率,平均每条记录的检测时间为毫秒级,适用于真实加密货币交易的检测场景。

关键词：隐私保护机器学习异常检测同态加密

基于Geohash的增强型位置k-匿名隐私保护方案

在线全文

学校读者我要写书评

暂无评论

计算机科学 2024年第9期51卷 393-400页

作者：李勇军祝跃飞白利芳信息工程大学网络空间安全学院郑州450007 中原工学院软件学院郑州450000 中国软件评测中心网络安全测评工程技术中心北京100048

随着LBS的广泛应用,位置隐私保护势在必行。近年来,作为应用较为广泛的位置k-匿名解决方案已成为研究热点,但k-匿名方案易受到敌手背景知识攻击,虽有学者们不同程度地考虑了位置相关的信息,但都不全面,并且当前形成匿名区的方案大多较... 详细信息

随着LBS的广泛应用,位置隐私保护势在必行。近年来,作为应用较为广泛的位置k-匿名解决方案已成为研究热点,但k-匿名方案易受到敌手背景知识攻击,虽有学者们不同程度地考虑了位置相关的信息,但都不全面,并且当前形成匿名区的方案大多较为耗时。基于此,为抵御敌手的语义攻击和查询及位置同质性攻击,提出了增强型位置k-匿名方案,在匿名区构建时充分考虑与物理位置相关的语义信息、时间属性、查询概率及查询语义等信息;然后在进行位置选取时,保证所选位置相对分散;为降低匿名区构建时耗,采用Geohash进行位置编码;最后通过真实数据集上的实验表明,所提方案可提供较好的位置隐私保护。

关键词： Geohash 增强型位置k-匿名基于位置的服务位置隐私位置语义查询概率时间属性

基于一种注意力机制U-Net的地震数据去噪方法

在线全文

学校读者我要写书评

暂无评论

石油地球物理勘探 2024年第4期59卷 724-735页

作者：曹静杰高康富许银坡王乃建张纯朱跃飞河北地质大学自然资源部京津冀城市群地下空间智能探测与装备重点实验室河北石家庄050031 河北地质大学河北省战略性关键矿产资源重点实验室河北石家庄050031 河北地质大学地球科学学院河北石家庄050031 东方地球物理公司采集技术中心河北涿州072751 中国矿业大学(北京)地球科学与测量工程学院北京100083

受野外采集过程中设备和环境等多种因素影响,地震数据中往往存在表面波、鬼波、随机噪声等各种噪声,影响了地震数据处理和解释的可靠性和准确性。近年来,基于人工智能的方法以其计算效率高、数值效果好等优点成为地震数据去噪的研究热点... 详细信息

受野外采集过程中设备和环境等多种因素影响,地震数据中往往存在表面波、鬼波、随机噪声等各种噪声,影响了地震数据处理和解释的可靠性和准确性。近年来,基于人工智能的方法以其计算效率高、数值效果好等优点成为地震数据去噪的研究热点。U型网络(U-Net)是一种经典的卷积神经网络结构,常用于图像分割任务;注意力机制(Attention Mechanism,AM)是一种能够让模型在学习过程中更加关注特定区域或特征的技术。通过在U-Net网络中添加AM模块,构建了一种具有注意力功能的U型网络(AU-Net),并将其运用到地震数据去噪。为解决去噪过程中产生的边界效应,使用膨胀填充的方法对数据进行切分,该方法通用性较高,可以用于其他网络模型。AU-Net和U-Net的去噪试验结果表明:AU-Net网络去噪的效果比U-Net更好,可更好地保留弱信号;同时,通过迁移学习使AU-Net去噪方法更具适应性。

关键词：地震勘探深度学习 U型网络地震数据去噪神经网络

Recent insights for the emerging COVID-19:Drug discovery,therapeutic options and vaccine development

同方期刊数据库博看期刊评论

在线全文

学校读者我要写书评

暂无评论

Asian Journal of Pharmaceutical Sciences 2021年第1期16卷 4-23页

作者： yuefei zhu Jia Li Zhiqing Pang Department of Biomedical Engineering Columbia UniversityNew York 10027USA Key Laboratory of Smart Drug Delivery Ministry of EducationSchool of PharmacyFudan UniversityShanghai 201203China Department of Biomedical Sciences Faculty of MedicineHealth and Human SciencesMacquarie UniversitySydney NSW 2109Australia

SARS-CoV-2 has been marked as a highly pathogenic coronavirus of COVID-19 disease into the human population,causing over 5.5 million confirmed cases *** COVID-19 has posed a global threat with significant human casualties and severe economic losses,there is a pressing demand to further understand the current situation and develop rational strategies to contain the drastic spread of the *** there are no specific antiviral therapies that have proven effective in randomized clinical trials,currently,the rapid detection technology along with several promising therapeutics for COVID-19 have mitigated its drastic ***,global institutions and corporations have commenced to parse out effective vaccines for the prevention of ***,the present review will give exhaustive details of extensive researches concerning the drug discovery and therapeutic options for COVID-19 as well as some insightful discussions of the status of COVID-19.

关键词： Coronavirus COVID-19 SARS-CoV-2 Precise prevention and control Drug discovery Vaccine development

An Efficient Authentication Infrastructure of WLAN Based on Self-Updating Hash Chain

在线全文

学校读者我要写书评

暂无评论

China Communications 2007年第1期4卷 66-74页

作者： Zhang, Haojun Chen, Li zhu, yuefei Henan Univ Technol Dept Comp Engn Chengchou Peoples R China

Two standards for security of WLAN,IEEE 802.11 i and WAPI,are introduced,and some shortcomings of them are pointed *** on hash chains technique to construct identity authentication tokens, a new authentication infrastructure for WLAN is proposed,in which AS authenticates the main entities -STAs in the foremost *** of the proposed scheme are discussed,including security, efficiency and feasibility *** with 802. 11 i and WAPI,the proposed scheme not only improves efficiency by reducing cost on computation and communication,but also promotes security property -availability by mitigating DoS attacks to AP and *** the same time,a self-updating hash chain technique is presented that is embedded in the proposed infrastructure to solve the length limitation of traditional hash chain.

关键词： WLAN authentication IEEE 802.11i WAPI self-updating hash chain

Detection of Malicious PDF Files Using a Two-Stage Machine Learning Algorithm

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2020年第6期29卷 1165-1177页

作者： HE Kang zhu yuefei HE Yubo LIU Long LU Bin LIN Wei State Key Laboratory of Mathematical Engineering and Advanced Computing

Portable document format(PDF) files are increasingly used to launch cyberattacks due to their popularity and increasing number of *** solutions have been developed to detect malicious files,but their accuracy decreases rapidly in face of new evasion *** explore how to improve the robustness of classifiers for detecting adversarial attacks in PDF *** replacement and the n-gram are implemented to extract robust features using proposed guiding *** the two-stage machine learning model,the objects are divided based on their types,and the anomaly detection model is first trained for each type *** former detection results are organized into tree-like information structure and treated as inputs to convolutional neural *** results show that the accuracy of our classifier is nearly 100% and the robustness against evasive samples is *** object features also enable the identification of different vulnerabilities exploited in malicious PDF files.

关键词： document handling adversarial attack detection Combinatorial mathematics robust feature extraction security of data Document processing techniques classifier robustness convolutional neural nets two-stage machine learning algorithm trees (mathematics) tree-like information structure Data security convolutional neural network feature extraction malicious files anomaly detection model two-stage machine learning model evasion techniques Neural computing techniques learning (artificial intelligence) pattern classification portable document format malicious PDF files

Architecture of Trusted PC

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Wuhan University Journal of Natural Sciences 2006年第6期11卷 1433-1436页

作者： WU Shuhua zhu yuefei Institute of Information Engineering Information Engineering University Zhengzhou 450002 Henan China

This paper, focusing on the trusted computing group＇s standards, explained the key concept of trusted compuling and provided the architecture of trusted PC. It built trust bottom-up by starting with trusted hardware and adding layers of trusted software. It is a system-level solution available to all applications running on the member platforms. This solution reduces the security burden on applications and thus simplifies application programming.

关键词： architecture trusted PC trusted platform