检索结果-内蒙古大学图书馆

4th International Conference on Advances in Computer Technology, Information Science and Communications, CTISC 2022

作者： Zheng, Liu Han, Qiu Junhu, Zhu Information & Engineering University State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China

ISBN: (数字)9781665458726

ISBN: (纸本)9781665458726

In the research of imbalanced data classification, the resampling effect of the existing resampling and random forest combination technology is greatly affected by the characteristic dimension of the training set, resulting in the unsatisfactory classification effect of the model. To solve this problem, a combination method of resampling and Random Forest for imbalanced data classification is proposed. The core idea is to resample the original data set multiple times by using the feature subset of each subtree in the random forest. By reducing the number of features used in a single resampling, the influence of feature dimension on resampling effect is reduced;Different feature subsets are used for resampling in different subtrees to improve the diversity of base classifiers. Based on SMOTE, B-SMOTE and ADASYN resampling techniques, three different optimization algorithms PSRF, PBSRF and PADRF are realized respectively. Taking the geometric mean and recall rate as evaluation indexes, the experimental results on 10 groups of KEEL data show that compared with the algorithm before optimization, the geometric mean values of the three algorithms are increased by 0.85%, 2.93% and 0.79% respectively, and the recall rates are increased by 1.44%, 3.26% and 1.07% respectively. © 2022 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Rotation Model Enhancement for Adversarial Attack

Rotation Model Enhancement for Adversarial Attack

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： Zhang, Hengwei Li, Zheming Liu, Hao Yang, Bo Li, Chenwei Wang, Jindong State Key Laboratory of Mathematical Engineering and Advanced Computing Henan Key Laboratory of Information Security Zhengzhou China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

Current white-box attack to deep neural networks have achieved considerable success, but not for black-box attack. The main reason is poor transferability, as the adversarial examples are crafted with single deep neural networks model, and excessively depend on that model. To address that problem, we propose a rotation model enhancement algorithm to craft adversarial examples. We improve rotation method in model enhancement. This algorithm constructs a possibility model to randomly rotate original images, and generates multiple transformed images. Therefore, we craft adversarial examples with single model, and boost attack on multiple models, which demonstrate considerable transferability and success rate for black-box attack. The simulation indicates the algorithm boost black-box attack with a 89.2% success rate. © 2022 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Misleading Image Classification with Multi-shearing and Random Padding 6

Misleading Image Classification with Multi-shearing and Rand...

引用

6th International Conference on Electronic Information Technology and Computer engineering, EITCE 2022

作者： Li, Chenwei Li, Zheming Yang, Bo Zhang, Hengwei State Key Laboratory of Mathematical Engineering and Advanced Computing Henan Key Laboratory of Information Security Zhengzhou China

ISBN: (纸本)9781450397148

Neural networks are vulnerable when input data is applied with human-imperceptible perturbations, which is called adversarial examples. When used in image classification models, adversarial examples mislead neural netwoks to classify images with wrong labels, posing great threat to network security. White-box attack has achieved considerable success rate, for the model structure is already known. But black-box attack remains to be improved, so as to the transferability. We refer to the model augmentation method in network training process, and apply to generating adversarial examples to reduce overfitting. Consulting fundamental methods in adversarial examples, we propose a multi-cropping transformation method to alleviate overfitting and enhance transferability. Firstly, referring to data augmentation, we multi-crop original images in every iteration with random possibilities in adversarial exaples generating process. Secondly, the gradient of model loss function is calculated, and the perturbations are added to original images. Finally, we generate adversarial examples with iterative perturbations. The validation of our method is verified on single models and ensemble models, and the transferability is improved, compared to other fundamental methods. © 2022 Association for computing Machinery.

关键词： Shearing

来源：评论

学校读者我要写书评

暂无评论

Code Clone Detection Based on Code Semantic Enhancement Model

Code Clone Detection Based on Code Semantic Enhancement Mode...

引用

2023 International Conference on Intelligent Communication and Computer engineering, ICICCE 2023

作者： Chunyan, Zhang Ziang, Shen Chen, Liang State Key of Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China Beijing University of Civil Engineering and Architecture Department of Electrical and Information Engineering Beijing China

ISBN: (纸本)9798350395617

Code clones are code fragments that are similar in composition or semantics of code statements, which can significantly improve software development efficiency and reduce costs by improving the quality of reusable code. However, existing code clone detection methods ignore the source code structure information and sequence information seriously. In this paper, we launched our research on the harder-to-break Type IV code clone detection, and propose a new method, namely Aug_Clone, to complete code clone detection task. Firstly, we utilize our published code semantic enhancement model to obtain the code semantic vector. Secondly, we perform code clone detection by calculating the similarity between these semantic vectors. At last, we evaluate our method on a large public data set, and experimental results show improvement over the state-of-the-art benchmarks, which demonstrates the effectiveness and efficiency of our method. © 2023 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Traceability Method of Network Attack Based on Evolutionary Game

Traceability Method of Network Attack Based on Evolutionary ...

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： Zhang, Hengwei Zhang, Xiaoning Sun, Pengyu Liu, Xiaohu Ma, Junqiang Zhang, Yuchen State Key Laboratory of Mathematical Engineering and Advanced Computing Henan Key Laboratory of Information Security Zhengzhou China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

Cyberspace is vulnerable to continuous malicious attacks. Traceability of network attacks is an effective defense means to curb and counter network attacks. In this paper, the evolutionary game model is used to analyze the network attack and defense behavior. On the basis of the quantification of attack and defense benefits, the replication dynamic learning mechanism is used to describe the change process of the selection probability of attack and defense strategies, and finally the evolutionary stability strategies and their solution curves of both sides are obtained. On this basis, the attack behavior is analyzed, and the probability curve of attack strategy and the optimal attack strategy are obtained, so as to realize the effective traceability of attack behavior. © 2022 IEEE.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

BContext2Name: Naming Functions in Stripped Binaries with Multi-Label Learning and Neural Networks 10

BContext2Name: Naming Functions in Stripped Binaries with Mu...

引用

10th IEEE International Conference on Cyber Security and Cloud computing and 9th IEEE International Conference on Edge computing and Scalable Cloud, CSCloud-EdgeCom 2023

作者： Xia, Bing Ge, Yunxiang Yang, Ruinan Yin, Jiabin Pang, Jianmin Tang, Chongjun Dept.ZhongYuan University of Technology Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China Dept.ZhongYuan University of Technology Henan Key Laboratory on Public Opinion Intelligent Analysis Zhengzhou China Dept.ZhongYuan University of Technology Zhengzhou China Dept.Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China

ISBN: (纸本)9798350312461

Conducting binary function naming helps reverse engineers understand the internal workings of the code and perform malicious code analysis without accessing the source code. However, the loss of debugging information poses the challenge of insufficient high-level semantic information description for stripping binary code function naming. Meanwhile, the existing binary function naming scheme has one function label for only one sample. The long-tail effect of function labels for a single sample makes the machine learning-based prediction models face the challenge. To obtain a function correlation label and improve the propensity score of uncommon tail labels, we propose a multi-label learning-based binary function naming model BContext2Name. This model automatically generates relevant labels for binary function naming by function context information with the help of PfastreXML model. The experimental results show that BContext2Name can enrich function labels and alleviate the long-tail effect that exists for a single sample class. To obtain high-level semantics of binary functions, we align pseudocode and basic blocks based on disassembly and decompilation, identify concrete or abstract values of API parameters by variable tracking, and construct API-enhanced control flow graphs. Finally, a seq2seq neural network translation model with attention mechanism is constructed between function multi-label learning and enhanced control flow graphs. Experiments on the dataset reveal that the F1 values of the BContext2Name model improve by 3.55% and 15.23% over the state-of-the-art XFL and Nero, respectively. This indicates that function multi-label learning can provide accurate labels for binary functions and can help reverse analysts understand the inner working mechanism of binary code. Code and data for this evaluation are available at https://***/CSecurityZhongYuan/BContext2Name. © 2023 IEEE.

关键词： Graphic methods

来源：评论

学校读者我要写书评

暂无评论

A Cross-Platform Instant Messaging User Association Method Based on Supervised Learning 4th

A Cross-Platform Instant Messaging User Association Method B...

引用

4th International Conference on Big Data and Security, ICBDS 2022

作者： Zhou, Pei Luo, Xiangyang Du, Shaoyong Shi, Wenqi Guo, Jiashan School of Cyber Science and Engineering Zhengzhou University Zhengzhou450000 China State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou450000 China

ISBN: (纸本)9789819932993

To solve the multi-platform user association problem of complex trajectory matching process and high time cost in cross-platform association positioning of instant messaging users, and at the same time make full use of the information in user trajectories, this paper proposes a supervised learning-based cross-platform instant messaging user association positioning method. The algorithm firstly places probes in the area where the target may appear to obtain user information;then gets user trajectories through the obtained user distance information and time information;selects user features through the classification algorithm of supervised learning, and designs a cross-platform instant messaging user association localization method based on supervised learning, so as to increase the association efficiency and accuracy of cross-platform instant messaging user association. The method conducts specific experiments for the most commonly used instant messaging tools in China, WeChat and Stranger users, and the results show that the method can achieve efficient and reliable association for these two types of instant messaging users. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

Social User Geolocation Method Based on POI Location Feature Enhancement in Context 1

引用

30th International Conference on Computational and Experimental engineering and Sciences, ICCES 2024

作者： Liu, Yu Qiao, Yaqiong Liu, Yimin Du, Shaoyong State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou China Key Laboratory of Cyberspace Situation Awareness of Henan Province Zhengzhou China School of Information Engineering North China University of Water Resources and Electric Power Zhengzhou China

ISBN: (数字)9783031687754

ISBN: (纸本)9783031687747

The geolocation information in social media user generated text is one of the important bases for inferring users' location. The existing methods often rely on the mentioned Point of Interest (POI) and location Indicative word in user generated text as the inference basis, in consideration of the lack of consideration of semantic information and the inaccuracy or incompleteness of POI contribute to the inaccuracy of user geolocation based on text. To address these issues, this paper proposes Social User Geolocation Method based on POI Location Feature Enhancement in Context (UGLE). Firstly, the POI in user-generated text is removed by deep neural network. Then, through three geographical traits of POI: location orientation, word frequency and word length, the problem of ambiguous POI location orientation in user generated text is solved. Finally, by combining the semantic information of the words around POI, the user vector with enhanced location features is generated, which improves the geolocation accuracy of social users. On the Weibo dataset, the proposed Method is compared with four existing typical methods, PaQL, SeFG, MNB-PART, and MNB-LGR. The results show that the proposed method's accuracy in inferring province-level and city-level locations is 76.78% and 62.56%, respectively. These results are 6.81% and 6.66% higher than the existing best results. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Research on Unpredetermined Behavior Recognition Techniques Based on Network Attacks 23

Research on Unpredetermined Behavior Recognition Techniques ...

引用

Proceedings of the 2023 5th International Conference on Internet of Things, Automation and Artificial Intelligence

作者： Yuxuan Xiao Jinlong Fei Junyi Wang State Key Laboratory of Mathematical Engineering and Advanced Computing China

ISBN: (纸本)9798400716485

With the increasing number of cyber-attack behaviors, it has caused great harm to the contemporary society. Currently, the cyber attack behavior recognition module is usually implemented by adopting a strategy based on rule-base matching, which often fails to identify the attack behaviors when faced with unpreset path attack behaviors of the actual attack process due to its reliance on predefined abnormal behavior patterns and attack labels. Therefore, this paper proposes an efficient and accurate unpreset behavior detection framework for cyber-attack behaviors, introduces machine learning techniques into the identification of unpreset attack behaviors, proposes an RF-Corr feature importance assessment method based on the Kendall correlation coefficient between features and the feature weight values, and designs an unpreset attack behavior identification method based on BiLSTM neural network, which improves the current effectiveness of identification and detection for unpresetted behavior of cyber attacks, and provides a solution for artificial intelligence detection of unpresetted behavior of cyber attacks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Solve the False-Positive Problem of the Mimic System Based on the Best Mimic Component Set

引用

China Communications 2022年第5期19卷 253-266页

作者： Yuwen Shao Zheng Zhang Xiaomei Wang Chuanxing Pan Jiangxing Wu State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou 450001China China National Digital Switching System Engineering and Technological R&D Center Zhengzhou 450002China

At present,there is a problem of false positives caused by the too vast mimic scope in mimic transformation *** studies have focused on the“compensation”method to deal with this problem,which is expensive and cannot fundamentally solve *** paper provides new insights into coping with the ***,this study summarizes the false-positive problem in the mimic transformation,analyzes its possible harm and the root ***,three properties about the mimic scope are *** on the three properties and security quantification technology,the best mimic component set theory is put forward to solve the false-positive *** are two algorithms,the supplemental method and the subtraction *** best mimic component set obtained by these two algorithms can fundamentally solve the mimic system’s false-positive problem but reduce the cost of mimic *** make up for the lack of previous researches.

关键词： mimic scope the best mimic component set false-positive rate mimic transformation

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：