检索结果-内蒙古大学图书馆

4th International Conference on Mechanical Design and Simulation, MDS 2024

作者： Zhou, Chen An, Baoran Yu, Bin Li, Suicheng Institute of Computer Application China Academy of Engineering Physics Mian Yang621900 China

ISBN: (纸本)9789819778867

In the practice of Model-Based Systems engineering (MBSE), where SysML models serve as the foundation for system integration frameworks, the maintenance of consistency of shared information across system models and other models requires the use of data exchange methods. A range of data exchange methods are proposed to address different exchange scenarios. This study surveyed literature related to these methods and elaborated on the three accomplished forms of data exchange and the concept of transformation based on the survey results. Subsequently, from the perspective of exchange characteristics, the literature concerning data exchange was categorized into five typical scenarios: model exchange, model transformation, model generation, code generation, and document generation. For each scenario, the research findings on relevant exchange methods were presented, and detailed descriptions of the advantages and disadvantages of some methods were provided. The results of this review can be used to guide the selection and improvement of data exchange methods. © The Author(s) 2025.

关键词： Data consistency

来源：评论

学校读者我要写书评

暂无评论

DLD: Defending Against Backdoors Using Differential Logits 6

DLD: Defending Against Backdoors Using Differential Logits

引用

6th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2024

作者： Yi, Ming Wang, Yulong Deng, Kai Yin, Mingyong Meng, Fanzhi Duan, Yubing China Academy of Engineering Physics Institute of Computer Application Mianyang621900 China

ISBN: (纸本)9798350316520

The existing research has shown that deep learning models are vulnerable to backdoor attacks, in which malicious actors can manipulate the model's predictions by embedding backdoors. As the stealthiness of these backdoor trigger mechanisms increases, traditional methods for detecting backdoors become increasingly inadequate. This paper presents a novel approach for detecting backdoors based on analyzing variations in model logits. Since the primary objective of a backdoor is to alter the model's prediction results, detecting backdoors through the analysis of logits variations offers a robust method for backdoor determination. To improve the accuracy of identifying backdoor models, our proposed method not only examines the logits variations of the model under investigation but also introduces an auxiliary normal model with a simpler structure. The presence of a backdoor in the model under investigation is determined by contrasting the differences in logits variations between the model under investigation and the normal model. Furthermore, to enhance detection speed, we introduce a dataset reduction mechanism to streamline the testing dataset. Experimental results demonstrate that our method successfully defends against high-stealth backdoor attacks such as WaNet and LIRA. Moreover, in terms of detection speed, our method outperforms mainstream methods like Neural Cleanse, which takes several hours for a single detection, by requiring only several minutes. © 2024 IEEE.

关键词： Statistical tests

来源：评论

学校读者我要写书评

暂无评论

A Random Walk-Based Approach for Clustering of Food Items 2nd

A Random Walk-Based Approach for Clustering of Food Items

引用

2nd International Conference on Cyber Intelligence and Information Retrieval, CIIR 2023

作者： Das, Priyanka Bhattacharya, Rupam Department of Computer Application Institute of Engineering and Management West Bengal Kolkata India

ISBN: (纸本)9789819735938

Most of the recipe sites have a distinct page structure comprising a title, list of ingredients, preparation instructions, comments, and thus it is very challenging to cluster similar recipes. The work presented here initially starts with the formation of a recipe network and finding ‘leading’ nodes from the network. To the best of our knowledge, there doesn’t exist any research that addresses the concept of finding ‘leading’ nodes from a recipe network and subsequently clustering similar recipes based on the ingredients. In order to fill this gap, we have proposed a method RWRC (Random Walk-based Recipe Clustering). To determine the ‘leading’ nodes, the proposed approach incorporates data from the breadth-first search tree. To create a breadth-first search (BFS) tree, the graph is explored in level-order, starting with each node as the root node. At each level of the traversal, we have computed the degree centrality, betweenness centrality and closeness centrality of each node. The ‘leading factor’ has been considered as the average of these centrality measures for all the nodes present in the network. A random walk-based graph clustering approach has been addressed that initiates from the root node of the original recipe network. If the ‘leading factor’ of the next visiting node in the random walk is equal to or greater than the average ‘leading factor’ of the whole network, the agent moves to its neighbours;otherwise, the stopping requirement is satisfied. Further, these nodes discovered by this random walk are organised into a cluster. When a walk comes to an end, the visited nodes from the initial graph are extracted, and the subsequent walk begins in the remaining graph. The random walk continues until there are no undiscovered node(s) left. These extracted nodes form the clusters of recipes. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Trees (mathematics)

来源：评论

学校读者我要写书评

暂无评论

A Crime Knowledge Discovery Scheme Based on Entity Recognition, Relation Extraction, and Development of Criminal Profiling Using Modus Operandi 11th

A Crime Knowledge Discovery Scheme Based on Entity Recognit...

引用

11th International Conference on Intelligent Computing and applications, ICRTC 2023

作者： Das, Priyanka Dhar, Sayan Banerjee, Shreedatri Basak, Shirsha Nandi, Sneha Saha, Sneha Department of Computer Application Institute of Engineering and Management West Bengal Kolkata India

ISBN: (纸本)9789819717231

In this paper, a criminal knowledge discovery scheme is proposed for analysing the crime situation in India. The approach pulls information regarding crime against women in Indian states and union territories from a sizable corpus of newspaper reports and then uses the retrieved corpus to identify named entities. India has experienced an increase in crime in recent years, and daily, several crime reports are published. However, it can be difficult to manually access the crime reports and extract relevant information that can aid in the analysis of crime trends by law enforcement. Even while there are many research on the analysis of crime patterns, relatively few of them include entity recognition, relation extraction, and features of modus operandi. This study focuses on phrase identification to discover the rules for determining modus operandi from police records. The effectiveness of the method is evaluated using precision and recall. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Crime

来源：评论

学校读者我要写书评

暂无评论

Pruning Optimization of Deep Learning Model Based on Generalization Estimation

Pruning Optimization of Deep Learning Model Based on General...

引用

2024 International Conference on Artificial Intelligence, Deep Learning and Neural Networks, AIDLNN 2024

作者： Li, Xuanzhe Ke, Wenjun Wen, Wantao Liu, Shaotong Sheng, Xiaojuan School of Computer Science and Engineering Southeast University China Beijing Institute of Computer Technology and Application China

ISBN: (纸本)9798331520816

Existing research has shown that deep learning models contain considerable redundancy, requiring compression optimization to eliminate unnecessary parameters and improve computational efficiency. In model compression through pruning, many existing methods often use loss values to guide the identification of sub-network models. However, low training loss does not necessarily indicate good generalization ability. Consequently, reducing model parameters while preserving strong generalization performance has become a key area of research. Addressing the challenge of ensuring strong generalization performance of submodels after optimization through deep learning model pruning, we propose a channel-structured generalization estimation model. The model architecture incorporates Fully Connected Neural Networks (FCNN) and Gated Recurrent Units (GRU) as modules for channel feature extraction. Additionally, we introduce a differentiable pruning algorithm based on generalization estimation. Experimental results demonstrate the capability of the channel-structured generalization estimation model in predicting the generalization gap via regression. Furthermore, the proposed differentiable pruning algorithm effectively safeguards the generalization capability of sub-models. We conduct pruning experiments on NiN-structured convolutional neural network models on two datasets, CIFAR-10 and SVHN. Comparative experimental results demonstrate the effectiveness of this approach: achieving a 0.72% accuracy improvement over the original model on CIFAR-10;and a 0.12% accuracy improvement over the base model on SVHN while reducing FLOPs by 50.2%, validating the effectiveness of the proposed approach. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

A Few-Shot Semantic Segmentation Network with Mixed Prototypes and Base Learner 7

A Few-Shot Semantic Segmentation Network with Mixed Prototyp...

引用

7th IEEE Advanced Information Technology, Electronic and Automation Control Conference, IAEAC 2024

作者： Wang, Han Meng, Ziyang Zhang, Yakun Institute of Computer Application China Academy of Engineering Physics Sichuan 621900 China

ISBN: (纸本)9798350339161

Few-shot semantic segmentation methods aim to segment new object categories with limited labeled samples. Existing approaches struggle to extract comprehensive semantic matching information from target regions and mitigate noise from complex backgrounds. To address background noise, this paper proposes a novel few-shot semantic segmentation approach using prototype generation and base class learning. A pyramid pooling convolutional layer extracts rich prototype vectors by discerning differences between the core mask and label mask. Additionally, a base class learning module captures semantic information from base class objects during training, effectively suppressing interference from background regions and aiding in background semantic matching. Comparative and ablation experiments demonstrate the superior performance of the proposed model. Evaluation on benchmark datasets shows its effectiveness, with results on Pascal-5i dataset surpassing the baseline model, achieving an average intersection over union of 67.15% in one-shot scenarios. © 2024 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Classification of Foods based on Bijective Soft Set 15

Classification of Foods based on Bijective Soft Set

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Das, Priyanka Bhattacharya, Rupam Dept. of Computer Application and Science Institute of Engineering & Management Kolkata India

ISBN: (纸本)9798350370249

Cuisine data is typically classified using images of the cuisine, cooking methods, cooking utensils, and, most importantly, ingredients. The vast amount of recipe data available on the internet necessitates the creation of specialised systems for storing, retrieving, and effectively utilising stored knowledge and data. Neural networks based expert systems are steadily being used to classify the cuisine data. But it lacks in dealing the vagueness present in the categorical cuisine data. Also, most of the popular recipes available online are mostly unhealthy. As a result, the Bijective soft set theory has been introduced into this study since it is an intelligent method for developing classification rules from food data and then recognising ambiguity and data reliance. The classification guidelines also address the necessity to select healthy/nutritious meals. The investigation outcomes from using bijective soft set analysis to a set of data samples are presented and appraised in this work. Furthermore, the classification rules derived from this approach provides valuable help for developing food recommendation systems. The accuracy of this classification scheme has got really influenced by application of the bijective soft set when compared to other state-of-the-art classifiers. ©2024 IEEE.

关键词： bijective soft set classification classification rule generation cuisine ingredients

来源：评论

学校读者我要写书评

暂无评论

Research on Program Change Impact Analysis Methods Based on Software Networks 24

Research on Program Change Impact Analysis Methods Based on ...

引用

24th IEEE International Conference on Software Quality, Reliability and Security Companion, QRS-C 2024

作者： Li, Shiyong Deng, Fei Guo, Xinglin Sun, Zhenxiao Zou, Zhuoliang Institute of Computer Application China Academy of Engineering Physics Sichuan Mianyang China

ISBN: (纸本)9798350365658

Aiming at the problem that the current program change impact analysis methods have coarse calculation granularity and without considering the actual situation of program change, results in overstating the impact propagation range and inability to quantify the degree of change impact, this paper proposes a program change impact analysis method based on the software network. Firstly, by combining the complex network theory to characterize the structural attributes and evolution patterns of software system, we abstract the software network morphology of system;Secondly, basing on the type of program change and the actual propagation of change impacts, two concepts of impact propagation decay rate and minimum threshold for impact marking are introduced to construct the simulation sub-model of impact propagation processes, enabling the analysis of the propagation range of program impacts;Thirdly, according to the structural properties of software network and the intrinsic attributes of nodes, the ripple degree simulation sub-model is constructed, and five quantization factors of the node's influenced degree are given to complete the quantitative assignment of the node's influenced degree;Finally, several open source software is used for application verification. Experimental results show that the proposed method aligns with reality in analyzing the propagation range of program change impacts, thus scientifically and reasonably reflecting the ripple effects of change impacts within the software system. At the same time, the method accurately and effectively quantifies the impact of program changes on other parts, thereby providing theoretical support for change impact analysis in software engineering. © 2024 IEEE.

关键词： change impact analysis impact propagation ripple effects software change software networks

来源：评论

学校读者我要写书评

暂无评论

Optimized graph transformer with molecule attention network based multi class attack detection framework for enhancing privacy and security in WSN

引用

Multimedia Tools and applications 2025年第15期84卷 14273-14304页

作者： Anoop, M. Mary, L. William Wilson, Allan J. Kiran, W.S. Department of Computer Science and Application SRM Institute of Science and Technology Ramapuram campus Chennai600089 India Department of Electronics and Communication Engineering Amrita College of Engineering and Technology Nagercoil India

In today's rapidly evolving network landscape, cybersecurity has become increasingly crucial. However, wireless sensor networks face unique challenges due to their limited resources and diverse composition, high computational demands, diminished accuracy rates, making them susceptible to a wide array of cyber threats. To tackle these challenges, this research introduced a novel method known as Optimized Graph Transformer with Molecule Attention Network (OGTMAN) on the fusion of Secure Multi-party Computation (SMC) and differential privacy aimed at enhancing privacy and security. This approach starts with data preprocessing, employing min–max normalization to address dataset imbalances. N-Tuple Contrastive Learning with Generative transformer(N-TCLG)feature extractor distinguish between different classes of data instances and effectively learns representations that capture relevant information about network traffic patterns Further, feature selection by enhanced chi-square statistics with kepler algorithm, is utilized to assess the significance of each feature in characteristic between normal network behavior and malicious activities further refines the feature selection method by optimizing the selection criteria based on the statistical properties of the dataset. Finally, the proposed OGTMAN classifies various attacks based on their behavioural patterns with the help of enhanced ant aphid optimization with adaptive step factor. Ultimately, this fine-tuning yield superior model training outcomes, resulting more accurate attack classification outcomes. To evaluate the effectiveness of this approach, datasets encompassing with KDD Cup 99 datasets are employed. The efficiency of the introduced method is demonstrated when compared with prevailing attack detection methods in terms of accuracy, recall, precision and F-measure of 99.98%, 99.5%.,99.9% and 99.5% respectively. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer N

关键词： Molecules

来源：评论

学校读者我要写书评

暂无评论

A Graph Based Approach For Detecting Sybil Node In Online Social Networks 15

A Graph Based Approach For Detecting Sybil Node In Online So...

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Chandra, Sucheta Das, Priyanka Institute of Engineering & Management Dept. of Computer Application and Science Kolkata India

ISBN: (纸本)9798350370249

Sybil attacks pose a significant threat to the security and integrity of online social networks by allowing malicious actors to create multiple fake identities to manipulate network behavior. The abstract outlines a method for detecting Sybil attacks in online social networks using graph analysis. It involves constructing a graph representation of the network, calculating node centrality measures such as in-degree and out-degree, and setting a degree threshold to identify potentially malicious nodes. Additional parameters like clustering coefficient and community structure are also considered to improve detection accuracy. The proposed approach aims to enhance the security of online social networks by identifying and isolating Sybil identities, thus preserving network integrity and trustworthiness. The authors' firmly assert that employing graph-based theory for Sybil node detection in online social networks stands out as one of the most effective methods. This approach not only offers robustness but also scalability, allowing for efficient detection even in large-scale networks. With its ability to unveil intricate relationships, anomalies, and patterns within the network, graph-based theory emerges as a cornerstone in the fight against Sybil attacks. © 2024 IEEE.

关键词： Cyber attacks

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：