检索结果-内蒙古大学图书馆

IoT Network Intrusion Detection Using Machine Learning Techniques

学校读者我要写书评

暂无评论

IoT Network Intrusion Detection Using Machine Learning Techn...

2022 International Conference on Smart Information systems and Technologies, SIST 2022

作者： Omarov, Bauyrzhan Auelbekov, Omirlan Koishiyeva, Tursynay Sadybekov, Ruslan Uxikbayev, Yerkebulan Bazarbayeva, Aizhan Al-Farabi Kazakh National University Department Of Inforamtion Systems Almaty Kazakhstan Institute Information And Computational Technologies Cs Mes Rk Almaty Kazakhstan Khoja Akhmet Yassawi International Kazakh-Turkish University Department Of Computer Engineering Turkestan Kazakhstan Khoja Akhmet Yassawi International Kazakh-Turkish University Department Of Computer Science Turkestan Kazakhstan

ISBN: (数字)9781665467902

ISBN: (纸本)9781665467902

The approaches that are directly suited for the implementation of the process of detecting anomalies in data are discussed in this article, and a comparison of the most widely used ones is carried out using the following criteria: the speed of implementation and the model's data needs. As a consequence, it can be stated that determining the quality of a model at the emission detection stage is challenging, and as a result, it must be examined at the final step (through subsequent analysis of the identified anomalies of various methods). Combinations of various strategies are the most often utilized and accurate procedures, as shown by experience. © 2022 IEEE.

关键词： Internet of things

CHGNN: A Semi-Supervised Contrastive Hypergraph Learning Network

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Song, Yumeng Gu, Yu Li, Tianyi Qi, Jianzhong Liu, Zhenghao Jensen, Christian S. Yu, Ge The School of Computer Science and Engineering Northeastern University Liaoning Shenyang110819 China The Department of Computer Science Aalborg University Aalborg9220 Denmark The School of Computing and Information Systems The University of Melbourne ParkvilleVIC3010 Australia

Hypergraphs can model higher-order relationships among data objects that are found in applications such as social networks and bioinformatics. However, recent studies on hypergraph learning that extend graph convolutional networks to hypergraphs cannot learn effectively from features of unlabeled data. To such learning, we propose a contrastive hypergraph neural network, CHGNN, that exploits self-supervised contrastive learning techniques to learn from labeled and unlabeled data. First, CHGNN includes an adaptive hypergraph view generator that adopts an auto-augmentation strategy and learns a perturbed probability distribution of minimal sufficient views. Second, CHGNN encompasses an improved hypergraph encoder that considers hyperedge homogeneity to fuse information effectively. Third, CHGNN is equipped with a joint loss function that combines a similarity loss for the view generator, a node classification loss, and a hyperedge homogeneity loss to inject supervision signals. It also includes basic and cross-validation contrastive losses, associated with an enhanced contrastive loss training process. Experimental results on nine real datasets offer insight into the effectiveness of CHGNN, showing that it outperforms 19 competitors in terms of classification accuracy consistently. Copyright © 2023, The Authors. All rights reserved.

关键词： Graph neural networks

Defining Problem from Solutions: Inverse Reinforcement Learning (IRL) and Its Applications for Next-Generation Networking

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Yinqiu Zhang, Ruichen Du, Hongyang Niyato, Dusit Kang, Jiawen Xiong, Zehui Kim, Dong In The School of Computer Science and Engineering Nanyang Technological University Singapore The School of Automation Guangdong University of Technology China The Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore The Department of Electrical and Computer Engineering Sungkyunkwan University Suwon16419 Korea Republic of

Performance optimization is a critical concern in networking, on which Deep Reinforcement Learning (DRL) has achieved great success. Nonetheless, DRL training relies on precisely defined reward functions, which formulate the optimization objective and indicate the positive/negative progress towards the optimal. With the ever-increasing environmental complexity and human participation in Next-Generation Networking (NGN), defining appropriate reward functions become challenging. In this article, we explore the applications of Inverse Reinforcement Learning (IRL) in NGN. Particularly, if DRL aims to find optimal solutions to the problem, IRL finds a problem from the optimal solutions, where the optimal solutions are collected from experts, and the problem is defined by reward inference. Specifically, we first formally introduce the IRL technique, including its fundamentals, workflow, and difference from DRL. Afterward, we present the motivations of IRL applications in NGN and survey existing studies. Furthermore, to demonstrate the process of applying IRL in NGN, we perform a case study about human-centric prompt engineering in Generative AI-enabled networks. We demonstrate the effectiveness of using both DRL and IRL techniques and prove the superiority of IRL. © 2024, CC BY.

关键词： Reinforcement learning

Single-Phase and Two-Phase Liquid Immersion Cooling of Data Center Power Supply Units for Heat Capture

学校读者我要写书评

暂无评论

Single-Phase and Two-Phase Liquid Immersion Cooling of Data ...

Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic systems (ITHERM)

作者： Haoyun Qiu Pouya Kabirzadeh Vivek S. Garimella Arjit Bali Keerthivasan Gurusami Kang Joon Lee Andrew Stillwell Todd Salamon Nenad Miljkovic Department of Mechanical Science and Engineering University of Illinois Urbana-Champaign Urbana IL Department of Electrical and Computer Engineering University of Illinois Urbana-Champaign Urbana IL Optical Systems and Device Research Laboratory Nokia Bell Labs Murray Hill NJ

ISBN: (数字)9798350364330

ISBN: (纸本)9798350364347

Data center power supply units (PSU) have high power density due to the increasing power demand and their low profiles. Air cooling proves to be increasingly difficult in mitigating the hot spots, and it also consumes extra energy in fluid delivery, air conditioning, and rejection to the ambience. In this work, we redesign and modify the PSU cooling infrastructure and adopt quiescent immersion cooling. This design removes all active air-cooling elements and enables heat capture. The PSU used in the experiments is a 2 kW 80 Plus Platinum commercial unit (Bel PET2000-12-074RA). The cooling performance was characterized by tracking the steady-state case temperatures of the switching semiconductors. Flutec PP1 was used to study the cooling performance variations with respect to PSU load and capture temperature. Conventional fan-cooled PSU with heat sinks served as the benchmark. Experiments show that two-phase immersion cooling grants an overall enhancement of 4.4× over the original air-cooling approach at 40 °C heat capture. The results suggest that heat rejection from the dielectric fluid to the riding heatsink is the main bottleneck, and further reduction of the thermal resistance would allow heat capture at a temperature closer to the liquid boiling point. The PSU thermal management approach studied here can help lower data center power usage effectiveness (PUE) while increasing PSU power delivery density.

关键词： Power system measurements Cooling Density measurement Thermal resistance Thermomechanical processes Immersion cooling Switches Thermal management Resistance heating Data center power

Unraveling the ‘Anomaly’ in Time Series Anomaly Detection: A Self-supervised Tri-domain Solution

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Sun, Yuting Pang, Guansong Ye, Guanhua Chen, Tong Hu, Xia Yin, Hongzhi School of Electrical Engineering and Computer Science The University of Queensland Australia School of Computing and Information Systems Singapore Management University Singapore Research and Development Department Deep Neural Computing Company Limited Department of Computer Science Rice University United States

The ongoing challenges in time series anomaly detection (TSAD), notably the scarcity of anomaly annotations and the variability in anomaly lengths and shapes, have led to the need for a more all-encompassing and efficient solution. As limited anomaly labels pose a significant obstacle for traditional supervised models in anomaly classification, various state-of-the-art deep learning techniques, such as self-supervised learning, have been introduced to tackle this issue. However, they encounter difficulties handling variations in anomaly lengths and shapes, limiting their adaptability to diverse anomalies. Additionally, many benchmark datasets suffer from the problem of having explicit anomalies that even random functions can detect. This problem is exacerbated by ill-posed evaluation metrics, such as point adjustment (PA) prior to scoring, which can result in inflated model performance. In this context, we propose a novel self-supervised learning based Tri-domain Anomaly Detector (TriAD), which addresses these challenges by modeling features across three data domains – temporal, frequency, and residual domains – without relying on anomaly labels. Unlike traditional contrastive learning methods, TriAD employs both inter-domain and intra-domain contrastive loss to learn common attributes among normal data and differentiate them from anomalies. Additionally, our approach can detect anomalies of varying lengths by integrating with a discord discovery algorithm. It is worth noting that this study is the first to reevaluate the deep learning potential in TSAD, utilizing both rigorously designed datasets (i.e., UCR Time Series Anomaly Archive) and evaluation metrics (i.e., PA%K and affiliation). Through experimental results on the UCR dataset, TriAD demonstrates a significant improvement in prediction performance across diverse evaluation metrics, achieving an impressive three-fold increase in PA%K based F1 scores over state-of-the-art (SOTA) deep learning models. Moreover

关键词： Anomaly detection

Space Shift Keying with Reconfigurable Intelligent Surfaces: Phase Configuration Designs and Performance Analysis

学校读者我要写书评

暂无评论

IEEE Open Journal of the Communications Society

IEEE Open Journal of the Communications Society 2021年 2卷 322-333页

作者： Li, Qiang Wen, Miaowen Wang, Shuai Alexandropoulos, George C. Wu, Yik-Chung Department of Electronic Engineering College of Information Science and Technology Jinan University Guangzhou510632 China School of Electronic and Information Engineering South China University of Technology Guangzhou510640 China Department of Electrical and Electronic Engineering Southern University of Science and Technology Shenzhen518055 China Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China Sifakis Research Institute of Trustworthy Autonomous Systems Shenzhen518055 China Department of Informatics and Telecommunications National and Kapodistrian University of Athens Athens15784 Greece Department of Electrical and Electronic Engineering University of Hong Kong Hong Kong Hong Kong

Reconfigurable intelligent surface (RIS)-assisted transmission and space shift keying (SSK) appear as promising candidates for future energy-efficient wireless systems. In this article, two RIS-based SSK schemes are proposed to efficiently improve the error and throughput performance of conventional SSK systems, respectively. The first one, termed RIS-SSK with passive beamforming (RIS-SSK-PB), employs an RIS for beamforming and targets the maximization of the minimum squared Euclidean distance between any two decision points. The second one, termed RIS-SSK with Alamouti space-time block coding (RIS-SSK-ASTBC), employs an RIS for ASTBC and enables the RIS to transmit its own Alamouti-coded information while reflecting the incident SSK signals to the destination. A low-complexity beamformer and an efficient maximum-likelihood (ML) detector are designed for RIS-SSK-PB and RIS-SSK-ASTBC, respectively. Approximate expressions for the average bit error probabilities of the source and/or the RIS are derived in closed-form assuming ML detection. Extensive computer simulations are conducted to verify the performance analysis. Results show that RIS-SSK-PB significantly outperforms the existing RIS-free and RIS-based SSK schemes, and RIS-SSK-ASTBC enables highly reliable transmission with throughput improvement. © 2020 IEEE.

关键词： Beamforming

A Survey on Recent Advances in Image Captioning

学校读者我要写书评

暂无评论

A Survey on Recent Advances in Image Captioning

2021 International Conference on Electrical, Electronics and Computing Technology, EECT 2021

作者： Chen, Feng Li, Xinyi Tang, Jintao Li, Shasha Wang, Ting Department of Computing Science College of Computer National University of Defense Technology China Science and Technology on Information Systems Engineering Laboratory National University of Defense Technology China

Image captioning, an interdisciplinary research field of computer vision and natural language processing, has attracted extensive attention. Image captioning aims to produce reasonable and accurate natural language sentences to describe images. It requires the captioning model to recognize objects and describe their relationships accurately. Intuitively, it is difficult for a machine to have the general image understanding ability like human beings. However, deep learning provides the basis for intelligent exploration. In this review, we will focus on recent advanced deep methods for image captioning. We classify existing methods into different categories and discuss these categories respectively;meanwhile, we discuss the related datasets and evaluation metrics. We also prospect the future research directions. © Published under licence by IOP Publishing Ltd.

关键词： Deep learning

DIFFUSION-BASED APPROACHES IN MEDICAL IMAGE GENERATION AND ANALYSIS

学校读者我要写书评

暂无评论

arXiv 2024年

作者： al Nomaan Nafi, Abdullah Hossain, Md. Alamgir Rifat, Rakib Hossain Uz Zaman, Md Mahabub Ahsan, Md Manjurul Raman, Shivakumar Department of Biomedical Engineering Islamic University Kushtia7003 Bangladesh Dept. of Computer Science Texas Tech University LubbockTX79409 United States Department of Industrial and Systems Engineering University of Oklahoma Norman Oklahoma73071 United States

Data scarcity in medical imaging poses significant challenges due to privacy concerns. Diffusion models, a recent generative modeling technique, offer a potential solution by generating synthetic and realistic data. However, questions remain about the performance of convolutional neural network (CNN) models on original and synthetic datasets. If diffusion-generated samples can help CNN models perform comparably to those trained on original datasets, reliance on patient-specific data for training CNNs might be reduced. In this study, we investigated the effectiveness of diffusion models for generating synthetic medical images to train CNNs in three domains: Brain Tumor MRI, Acute Lymphoblastic Leukemia (ALL), and SARS-CoV-2 CT scans. A diffusion model was trained to generate synthetic datasets for each domain. Pre-trained CNN architectures were then trained on these synthetic datasets and evaluated on unseen real data. All three datasets achieved promising classification performance using CNNs trained on synthetic data. Local Interpretable Model-Agnostic Explanations (LIME) analysis revealed that the models focused on relevant image features for classification. This study demonstrates the potential of diffusion models to generate synthetic medical images for training CNNs in medical image analysis. © 2024, CC BY.

关键词： Convolutional neural networks

Data-driven Optimal Cooperative Output Regulation of Unknown Linear Discrete-time Multi-agent systems

学校读者我要写书评

暂无评论

Data-driven Optimal Cooperative Output Regulation of Unknown...

Chinese Control and Decision Conference, CCDC

作者： Dong Liang Yi Dong Chaoli Wang Engang Tian Department of Control Science and Engineering School of Optical-Electrical and Computer Engineering University of Shanghai for Science and Technology Shanghai China College of Electronic and Information Engineering Shanghai Research Institute for Intelligent Autonomous Systems Tongji University Shanghai China

The cooperative output regulation problem has been extensively studied on the basis of the distributed observer approach. However, the majority of the existing research assumes that the dynamics is known previously. To remove this condition, the cooperative output regulation problem is further solved via the data-driven framework where the dynamics of the plant is unknown. First, a data-driven distributed observer is established to estimate the state of the leader with unknown dynamics subject to external inputs. Second, the unknown regulator equations are solved using the iterative recurrent neural network approach. Third, the state-based data-driven distributed control law is synthesized to solve the problem. The optimal gains are derived by solving convex optimization problems using input and state data. Finally, a numerical example is presented to verify the feasibility of the proposed framework.

关键词：