检索结果-内蒙古大学图书馆

GAT-COBO: Cost-Sensitive Graph Neural Network for Telecom Fraud Detection

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Hu, Xinxin Chen, Haotian Zhang, Junjie Chen, Hongchang Liu, Shuxin Li, Xing Wang, Yahui Xue, Xiangyang National Digital Switching System Engineering and Technological Research Center Zhengzhou450002 China Department of Electrical & Computer Engineering University of Toronto TorontoM5S 3G4 Canada Institute of Big Data Fudan University Shanghai China

Along with the rapid evolution of mobile communication technologies, such as 5G, there has been a drastically increase in telecom fraud, which significantly dissipates individual fortune and social wealth. In recent years, graph mining techniques are gradually becoming a mainstream solution for detecting telecom fraud. However, the graph imbalance problem, caused by the Pareto principle, brings severe challenges to graph data mining. This is a new and challenging problem, but little previous work has been noticed. In this paper, we propose a Graph ATtention network with COst-sensitive BOosting (GAT-COBO) for the graph imbalance problem. First, we design a GAT-based base classifier to learn the embeddings of all nodes in the graph. Then, we feed the embeddings into a well-designed cost-sensitive learner for imbalanced learning. Next, we update the weights according to the misclassification cost to make the model focus more on the minority class. Finally, we sum the node embeddings obtained by multiple cost-sensitive learners to obtain a comprehensive node representation, which is used for the downstream anomaly detection task. Extensive experiments on two real-world telecom fraud detection datasets demonstrate that our proposed method is effective for the graph imbalance problem, outperforming the state-of-the-art GNNs and GNN-based fraud detectors. In addition, our model is also helpful for solving the widespread over-smoothing problem in GNNs. The GAT-COBO code and datasets are available at https://***/xxhu94/GAT-COBO. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

RJSSIM: A reconfigurable job scheduling smulator for parallel processing learning

学校读者我要写书评

暂无评论

RJSSIM: A reconfigurable job scheduling smulator for paralle...

Frontiers in Education (FIE) Conference

作者： L.FW. Goes C.A.P.S. Martins Electrical Engineering Computational and Digita Systems Laboratory Pontifical Catholic University of Minas Gerais Belo Horizonte Minas Gerais Brazil Computer Science Department Elechical Engineering Computational and Digital System Laboratory Pontifical Catholic University of Minas Gerais Belo Horizonte Minas Gerais Brazil

A joint energy and latency framework for transfer learning over 5G industrial edge networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yang, Bo Fagbohungbe, Omobayode Cao, Xuelin Yuen, Chau Qian, Lijun Niyato, Dusit Zhang, Yan Engineering Product Development Pillar Singapore University of Technology and Design Singapore487372 Singapore Department of Electrical and Computer Engineering Credit Center Prairie View A and M University Texas A and M University System Prairie ViewTX77446 United States School of Computer Science and Engineering Nanyang Technological University Singapore Singapore University of Oslo Norway Simula Metropolitan Center for Digital Engineering Norway

In this paper, we propose a transfer learning (TL)-enabled edge-CNN framework for 5G industrial edge networks with privacy-preserving characteristic. In particular, the edge server can use the existing image dataset to train the CNN in advance, which is further fine-tuned based on the limited datasets uploaded from the devices. With the aid of TL, the devices that are not participating in the training only need to fine-tune the trained edge-CNN model without training from scratch. Due to the energy budget of the devices and the limited communication bandwidth, a joint energy and latency problem is formulated, which is solved by decomposing the original problem into an uploading decision subproblem and a wireless bandwidth allocation subproblem. Experiments using ImageNet demonstrate that the proposed TL-enabled edge-CNN framework can achieve almost 85% prediction accuracy of the baseline by uploading only about 1% model parameters, for a compression ratio of 32 of the autoencoder. Copyright © 2021, The Authors. All rights reserved.

关键词： Convolutional neural networks

Federated Cubic Regularized Newton Learning with Sparsification-amplified Differential Privacy

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Huo, Wei Liu, Changxin Ding, Kemi Johansson, Karl Henrik Shi, Ling Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Hong Kong Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Sweden Digital Futures StockholmSE-10044 Sweden School of System Design and Intelligent Manufacturing Southern University of Science and Technology Shenzhen518055 China

This paper investigates the use of the cubic-regularized Newton method within a federated learning framework while addressing two major concerns that commonly arise in federated learning: privacy leakage and communication bottleneck. We introduce a federated learning algorithm called Differentially Private Federated Cubic Regularized Newton (DP-FCRN). By leveraging second-order techniques, our algorithm achieves lower iteration complexity compared to first-order methods. We also incorporate noise perturbation during local computations to ensure privacy. Furthermore, we employ sparsification in uplink transmission, which not only reduces the communication costs but also amplifies the privacy guarantee. Specifically, this approach reduces the necessary noise intensity without compromising privacy protection. We analyze the convergence properties of our algorithm and establish the privacy guarantee. Finally, we validate the effectiveness of the proposed algorithm through experiments on a benchmark dataset. © 2024, CC BY.

关键词： Federated learning

Considering Multi-Scale Data for Continuous Traffic Prediction Using Adaptive Multi-Agent system

学校读者我要写书评

暂无评论

Considering Multi-Scale Data for Continuous Traffic Predicti...

International Conference on Intelligent Transportation

作者： Ha Nhi Ngo Elsy Kaddoum Matej Cabecauer Erik Jenelius Anaïs Goursolle Continental Digital Services France Toulouse France Cooperative Multi-Agent System team IRIT Computer Science Research Institute of Toulouse Paul Sabatier University Toulouse France Department of Civil and Architectural Engineering KTH Royal Institute of Technology Stockholm Sweden data scientist at Continental Digital Service France Toulouse France

Accurate traffic prediction is essential for effective traffic management and planning. However, traffic prediction models are challenged by various factors such as complex spatiotemporal dependencies in traffic data. Recently, researchers have explored the new approach known as stream analysis since it can continuously update models by capturing new behaviors from the traffic data stream. However, applying this approach specifically raises the question about the balance between model complexity and model flexibility for dynamic updates. ADRIP - Adaptive multi-agent system for DRIving behaviors Prediction proposed in [1], [2] has combined the dynamic clustering and the multi-agent system approach to solve this challenge. This system has been applied to predict traffic dynamics at the road segment level. In this paper, we aim to extend ADRIP to complete its functionality for traffic prediction at the network level. Experiments for multi-scale traffic data are conducted to compare extended ADRIP with well-known clustering models.

关键词：

Trend and Correlation Analysis of Instagram Activity Using Data Mining and Statistics

学校读者我要写书评

暂无评论

Trend and Correlation Analysis of Instagram Activity Using D...

International Conference on Cybernetics and Intelligent system (ICORIS)

作者： Wilem Musu Indra Samsie Aldi Bastiatul Fawait Nadia Lempan Nurliah Maechel Maximiliano Gabriel Sinar Department of Informatics Engineering Dipa Makassar University Makassar Indonesia Department of Information System Dipa Makassar University Makassar Indonesia Department of Computer Science Widya Gama Mahakan University Samarinda Indonesia Department of Digital Business Dipa Makassar University Makassar Indonesia Department of Entrepreneurship Dipa Makassar University Makassar Indonesia

ISBN: (数字)9798331539603

ISBN: (纸本)9798331539610

With the growing importance of social media for brand engagement, there remains a limited understanding of how specific metrics-such as Reach, Followers, and New Followers-impact Instagram user interaction over time. This research seeks to address this gap by examining which variables most effectively drive Engagement on Instagram, providing insights for strategic content optimization. Through an analysis of 100 posts, using multiple linear regression and decision tree methods, the study investigates the relationship between Reach, Followers, New Followers, and Engagement. Results indicate that Reach, Followers, and New Followers are not statistically significant predictors of Engagement, as evidenced by high p-values (all> 0.05). Decision tree results demonstrate a Mean Absolute Error (MAE) of 0.0226 and an $\mathrm{R}^{2}$ Score of 0.8003, suggesting moderate predictive accuracy but with potential for refinement. Moving average analysis reveals fluctuating Engagement with no stable long-term trend. A correlation matrix further shows that Reach and New Followers exhibit high correlations with Engagement (0.9723 and 0.9707, respectively), while Followers show a weak correlation (0.2575). These findings highlight the limited impact of Followers on Engagement, underscoring the importance of Reach and New Followers. This study contributes novel insights by identifying key engagement drivers, emphasizing the need for more targeted strategies in Instagram campaign planning to enhance user interaction.

关键词： Correlation Social networking (online) Linear regression Market research Timing Planning Web sites Multimedia communication Decision trees Optimization

Optimization on multi-user physical layer security of intelligent reflecting surface-aided VLC

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Sun, Shiyuan Yang, Fang Song, Jian Han, Zhu Department Of Electronic Engineering Beijing National Research Center For Information Science And Technology Tsinghua University Beijing100084 China Key Laboratory Of Digital TV System Of Guangdong Province And Shenzhen City Research Institute Of Tsinghua University In Shenzhen Shenzhen518057 China Department Of Electrical And Computer Engineering The University Of Houston HoustonTX77004 United States Department Of Computer Science And Engineering Kyung Hee University Seoul446-701 Korea Republic of

This letter investigates physical layer security in intelligent reflecting surface (IRS)-aided visible light communication (VLC). Under the point source assumption, we first elaborate the system model in the scenario with multiple legitimate users and one eavesdropper, where the secrecy rate maximization problem is transformed into an assignment problem by objective function approximation. Then, an iterative Kuhn-Munkres algorithm is proposed to optimize the transformed problem, and its computational complexity is in the second-order form of the numbers of IRS units and transmitters. Moreover, numerical simulations are carried out to verify the approximation performance and the VLC secrecy rate improvement by IRS. Copyright © 2022, The Authors. All rights reserved.

关键词： Combinatorial optimization

Optical Integrated Sensing and Communication: Architectures, Potentials and Challenges

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wen, Yunfeng Yang, Fang Song, Jian Han, Zhu The Department of Electronic Engineering Tsinghua University Beijing100084 China The Key Laboratory of Digital TV System of Shenzhen City Research Institute of Tsinghua University in Shenzhen Shenzhen518057 China The Shenzhen International Graduate School Tsinghua University Shenzhen518055 China The Department of Electrical and Computer Engineering University of Houston HoustonTX77004 United States The Department of Computer Science and Engineering Kyung Hee University Seoul446-701 Korea Republic of

Integrated sensing and communication (ISAC) is viewed as a crucial component of future mobile networks and has gained much interest in both academia and industry. Similar to the emergence of radio-frequency (RF) ISAC, the integration of free space optical communication and optical sensing yields optical ISAC (O-ISAC), which is regarded as a powerful complement to its RF counterpart. In this article, we first introduce the generalized system structure of O-ISAC, and then elaborate on three advantages of O-ISAC, i.e., increasing communication rate, enhancing sensing precision, and reducing interference. Next, waveform design and resource allocation of O-ISAC are discussed based on pulsed waveform, constant-modulus waveform, and multi-carrier waveform. Furthermore, we put forward future trends and challenges of O-ISAC, which are expected to provide some valuable directions for future research. Copyright © 2023, The Authors. All rights reserved.

关键词： Optical communication

Free Space Optical Integrated Sensing and Communication Based on DCO-OFDM: Performance Metrics and Resource Allocation

学校读者我要写书评

暂无评论

arXiv 2023年

As one of the six usage scenarios of the sixth generation (6G) mobile communication system, integrated sensing and communication (ISAC) has garnered considerable attention, and numerous studies have been conducted on radio-frequency (RF)-ISAC. Benefitting from the communication and sensing capabilities of an optical system, free space optical (FSO)-ISAC becomes a potential complement to RF-ISAC. In this paper, a direct-current-biased optical orthogonal frequency division multiplexing (DCO-OFDM) scheme is proposed for FSO-ISAC. To derive the spectral efficiency for communication and the Fisher information for sensing as performance metrics, we model the clipping noise of DCO-OFDM as additive colored Gaussian noise to obtain the expression of the signal-to-noise ratio. Based on the derived performance metrics, joint power allocation problems are formulated for both communication-centric and sensing-centric scenarios. In addition, the non-convex joint optimization problems are decomposed into sub-problems for DC bias and subcarriers, which can be solved by block coordinate descent algorithms. Furthermore, numerical simulations demonstrate the proposed algorithms and reveal the trade-off between communication and sensing functionalities of the OFDM-based FSO-ISAC system. Copyright © 2023, The Authors. All rights reserved.

关键词： Orthogonal frequency division multiplexing