检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Dwork, Cynthia Hays, Chris Kleinberg, Jon Raghavan, Manish Department of Computer Science Harvard University United States Institute for Data Systems and Society Massachusetts Institute of Technology United States Departments of Computer Science and Information Science Cornell University United States Sloan School of Management Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology United States

We study the impact of content moderation policies in online communities. In our theoretical model, a platform chooses a content moderation policy and individuals choose whether or not to participate in the community according to the fraction of user content that aligns with their preferences. The effects of content moderation, at first blush, might seem obvious: it restricts speech on a platform. However, when user participation decisions are taken into account, its effects can be more subtle - and counter-intuitive. For example, our model can straightforwardly demonstrate how moderation policies may increase participation and diversify content available on the platform. In our analysis, we explore a rich set of interconnected phenomena related to content moderation in online communities. We first characterize the effectiveness of a natural class of moderation policies for creating and sustaining communities. Building on this, we explore how resource-limited or ideological platforms might set policies, how communities are affected by differing levels of personalization, and competition between platforms. Our model provides a vocabulary and mathematically tractable framework for analyzing platform decisions about content moderation. Copyright © 2023, The Authors. All rights reserved.

关键词： Online systems

来源：评论

学校读者我要写书评

暂无评论

Overview and Assessment Approaches of Energy Consumption and GHG Emissions Through the Development and Application of Artificial Intelligence - An Example from the Agricultural Sector

Overview and Assessment Approaches of Energy Consumption and...

引用

2024 IEEE International Workshop on Metrology for Agriculture and Forestry, MetroAgriFor 2024

作者： Rilling, Stefan Henningsen, Jens Milde, Katharina Martini, Daniel Wickert, Lorenz Kelbert, Patricia Fraunhofer Institute for Intelligent Analysis and Information Systems Iais Adaptive Reflective Teams Sankt Augustin Germany Fraunhofer Institute for Experimental Software Engineering Iese Data Science Kaiserslautern Germany Advisory Board for Technology and Construction in Agriculture Ktbl Team Digital Technologies Darmstadt Germany

ISBN: (纸本)9798350355444

This article looks at the sustainability of various artificial intelligence (AI) methods. In recent months, AI methods have received a great deal of public attention and the benefits for sustainability have been emphasised. However, the ecological impacts of AI methods must also be considered in this debate. This article presents an overview of various approaches to sustainability assessment and addresses issues relating to energy consumption and the associated Carbon dioxide (CO2) emissions, and a Life Cycle Assessment (LCA) calculation for a concrete agricultural AI-use case. © 2024 IEEE.

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

Training time minimization for federated edge learning with optimized gradient quantization and bandwidth allocation

引用

Frontiers of information Technology & Electronic engineering 2022年第8期23卷 1247-1263页

作者： Peixi LIU Jiamo JIANG Guangxu ZHU Lei CHENG Wei JIANG Wu LUO Ying DU Zhiqin WANG State Key Laboratory of Advanced Optical Communication Systems and Networks Department of ElectronicsPeking UniversityBeijing 100871China China Academy of Information and Communications Technology Beijing 100191China Shenzhen Research Institute of Big Data Shenzhen 518172China College of Information Science and Electronic Engineering Zhejiang UniversityHangzhou 310027China Zhejiang Provincial Key Laboratory of Information Processing Communication and NetworkingHangzhou 310027China

Training a machine learning model with federated edge learning(FEEL)is typically time consuming due to the constrained computation power of edge devices and the limited wireless resources in edge *** this study,the training time minimization problem is investigated in a quantized FEEL system,where heterogeneous edge devices send quantized gradients to the edge server via orthogonal *** particular,a stochastic quantization scheme is adopted for compression of uploaded gradients,which can reduce the burden of per-round communication but may come at the cost of increasing the number of communication *** training time is modeled by taking into account the communication time,computation time,and the number of communication *** on the proposed training time model,the intrinsic trade-off between the number of communication rounds and per-round latency is ***,we analyze the convergence behavior of the quantized FEEL in terms of the optimality ***,a joint data-and-model-driven fitting method is proposed to obtain the exact optimality gap,based on which the closed-form expressions for the number of communication rounds and the total training time are *** by the total bandwidth,the training time minimization problem is formulated as a joint quantization level and bandwidth allocation optimization *** this end,an algorithm based on alternating optimization is proposed,which alternatively solves the subproblem of quantization optimization through successive convex approximation and the subproblem of bandwidth allocation by bisection *** different learning tasks and models,the validation of our analysis and the near-optimal performance of the proposed optimization algorithm are demonstrated by the simulation results.

关键词： Federated edge learning Quantization optimization Bandwith allocation Training time minimization

来源：评论

学校读者我要写书评

暂无评论

Copula Guided Parallel Gibbs Sampling for Nonparametric and Coherent Topic Discovery (Extended Abstract) 39

Copula Guided Parallel Gibbs Sampling for Nonparametric and ...

引用

39th IEEE International Conference on data engineering, ICDE 2023

作者： Lin, Lihui Rao, Yanghui Xie, Haoran Lau, Raymond Y. K. Yin, Jian Wang, Fu Lee Li, Qing Sun Yat-sen University School of Computer Science and Engineering Guangzhou China Lingnan University Department of Computing and Decision Sciences New Territories Hong Kong City University of Hong Kong Department of Information Systems Kowloon Hong Kong Sun Yat-sen University Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Hong Kong Metropolitan University School of Science and Technology Kowloon Hong Kong The Hong Kong Polytechnic University Department of Computing Kowloon Hong Kong

ISBN: (纸本)9798350322279

In terms of the generative process, the Gamma-Gamma-Poisson Process (G2PP) is equivalent to the nonparametric topic model of Hierarchical Dirichlet Process (HDP). Considering the high computational cost of estimating parameters in HDP, a parallel G2PP was developed to generate topics efficiently via multi-threading. Unfortunately, the above model needs to predefine the number of topics. To address this issue, we first propose a Topic Self-Adaptive Model (TSAM) for nonparametric and parallel topic discovery. In TSAM, a monitor-executor mechanism is developed to manage the global topic information using a hierarchical structure of threads. Based on the apparatus of copulas, we further extend our TSAM to TSAMcop for coherent topic modeling by exploiting a copula guided parallel Gibbs sampling algorithm. Extensive experiments validate the effectiveness of both TSAM and TSAMcop. © 2023 IEEE.

关键词： information management

来源：评论

学校读者我要写书评

暂无评论

Instrumental variable value iteration for causal offline reinforcement learning

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2024年第1期25卷 14624-14679页

作者： Luofeng Liao Zuyue Fu Zhuoran Yang Yixin Wang Dingli Ma Mladen Kolar Zhaoran Wang Department of Industrial Engineering and Operations Research Columbia University New York NY Department of Industrial Engineering and Management Sciences Northwestern University Evanston IL Department of Statistics and Data Science Yale University New Haven CT Department of Statistics University of Michigan Ann Arbor MI Department of Information Systems and Operations Management Michael G. Foster School of Business University of Washington Seattle WA epartment of Data Sciences and Operations Marshall School of Business University of Southern California Los Angeles CA

In offline reinforcement learning (RL) an optimal policy is learned solely from a priori collected observational data. However, in observational data, actions are often confounded by unobserved variables. Instrumental variables (IVs), in the context of RL, are the variables whose influence on the state variables is all mediated by the action. When a valid instrument is present, we can recover the confounded transition dynamics through observational data. We study a confounded Markov decision process where the transition dynamics admit an additive nonlinear functional form. Using IVs, we derive a conditional moment restriction through which we can identify transition dynamics based on observational data. We propose a provably efficient IV-aided Value Iteration (IVVI) algorithm based on a primal-dual reformulation of the conditional moment restriction. To our knowledge, this is the first provably efficient algorithm for instrument-aided offline RL.

关键词： instrumental variables reinforcement learning causal inference

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Micro-Segmentations for Zero-Trust Services via Large Language Model (LLM)-enhanced Graph Diffusion

arXiv

引用

arXiv 2024年

作者： Liu, Yinqiu Liu, Guangyuan Du, Hongyang Niyato, Dusit Kang, Jiawen Xiong, Zehui In Kim, Dong Shen, Xuemin College of Computing and Data Science Nanyang Technological University Singapore School of Automation Guangdong University of Technology China Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore Department of Electrical and Computer Engineering Sungkyunkwan University Korea Republic of Department of Electrical and Computer Engineering University of Waterloo Canada

—In the rapidly evolving Next-Generation Networking (NGN) era, the adoption of zero-trust architectures has become increasingly crucial to protect security. However, provisioning zero-trust services in NGNs poses significant challenges, primarily due to the environmental complexity and dynamics. Motivated by these challenges, this paper explores efficient zero-trust service provisioning using hierarchical micro-segmentations. Specifically, we model zero-trust networks via hierarchical graphs, thereby jointly considering the resource- and trust-level features to optimize service efficiency. We organize such zero-trust networks through micro-segmentations, which support granular zero-trust policies efficiently. To generate the optimal micro-segmentation, we present the Large Language Model-Enhanced Graph Diffusion (LEGD) algorithm, which leverages the diffusion process to realize a high-quality generation paradigm. Additionally, we utilize policy boosting and Large Language Models (LLM) to enable LEGD to optimize the generation policy and understand complicated graphical features. Moreover, realizing the unique trustworthiness updates or service upgrades in zero-trust NGN, we further present LEGD-Adaptive Maintenance (LEGD-AM), providing an adaptive way to perform task-oriented fine-tuning on LEGD. Extensive experiments demonstrate that the proposed LEGD achieves 90% higher efficiency in provisioning services compared with other baselines. Moreover, the LEGD-AM can reduce the service outage time by over 50%. © 2024, CC BY.

关键词： Diffusion

来源：评论

学校读者我要写书评

暂无评论

Adaptive Student Inference Network for Efficient Single Image Super-Resolution

Adaptive Student Inference Network for Efficient Single Imag...

引用

IEEE International Conference on data Mining (ICDM)

作者： Kang Miao Zhao Zhang Jiahuan Ren Mingbo Zhao Haijun Zhang Richang Hong School of Computer Science and Information Engineering Hefei University of Technology Hefei China Key Laboratory of Knowledge Engineering with Big Data (Ministry of Education) & Intelligent Interconnected Systems Laboratory of Anhui Province Hefei University of Technology Hefei China Yunnan Key Laboratory of Software Engineering Kunming China School of Information Science and Technology Donghua University Shanghai China Department of Computer Science Harbin Institute of Technology (Shenzhen) Shenzhen China

Recent advances in single image super-resolution (SISR) have achieved remarkable performance through deep learning. However, the high computational cost hinders the deployment of SISR models on edge devices. Instead of proposing new SISR models, a new trend is emerging to improve network efficiency by reducing parameters, FLOPs, and inference time through slight modifications to the original models. However, recent methods usually focus on reducing only one of three metrics, i.e., FLOPs, parameters and inference time, which inevitably increases the other two metrics. In this paper, we propose a novel Adaptive Student Inference Network (ASIN) on popular SISR models, which aims at reducing FLOPs and inference time while maintaining the number of parameters and restoring clearer high-resolution images. Specifically, our ASIN divides a SISR model into three components (head, body and tail) and adopts various strategies for each part. For head and tail parts, to ensure the restored images contain more detailed information, a novel auxiliary Enhanced Teacher Network (ETNet) is designed, which is trained with the ground-truth images to obtain more prior knowledge to guide student network to extract more accurate textures using a new knowledge distillation method. For the body part, owing to the varying difficulties of the reconstructions in different regions, we propose an Adaptive Depth Predicted Module (ADPM) to dynamically shorten average depth of network to reduce the computational cost of overall network. Extensive experiments on two datasets demonstrate the effectiveness and state-of-the-art performance of our ASIN compared to its counterparts.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Advanced Malware Classification Using Convolutional Neural Networks and Image Transformation Techniques

Advanced Malware Classification Using Convolutional Neural N...

引用

Advancements and Key Challenges in Green Energy and Computing (AKGEC), International Conference on

作者： Shweta Bhattacharjee Porna Md Zakir Hossain Abdullah Al Naseeh Chowdhury M Govind Raj Roise Uddin Mukit Al Elahi Abdul Basit Mohammad Aman Ullah Khan Girigula Durga Bhavani Department of Computer Science and Engineering Ahsanullah University of Science and Technology Dhaka Bangladesh Department of Data Science Grand Canyon University USA Computer Science and Engineering Sylhet Engineering College Sylhet Bangladesh Department of ECE Malla Reddy Engineering College Maisammaguda Telangana India Computer and Information Systems Security Cyber California State University San Bernardino USA Department of CSE Pacific states university Los Angeles California USA Department of CSE Pacific states university 4651 Rosewood St USA Department of CSE School of Data and Sciences BRAC University Dhaka Bangladesh School of Computer Science and Engineering VIT-AP University Amaravati Andhra Pradesh India

ISBN: (数字)9798350390179

ISBN: (纸本)9798350390186

Malware seriously compromises cybersecurity and calls for sophisticated and effective categorization methods to find and lessen its effects. This research article offers a complete method based on image processing and deep learning to classify malware. ResNet18, VGG16, VGG19, AlexNet, and our custom-designed MalwareNet are among the several deep learning models we use to categorize malware photos from the nine different classes in the Big 2015 dataset-which shows a clear class imbalance. Using hyperband and bayesian optimization approaches, our strategy maximizes hyperparameters to provide the greatest performance. The hyperparameters for ResNet18, VGG16, VGG19, and AlexNet models comprise learning rate, kernel size, number of epochs, dropout rate, and dense layer size as well as particular parameters for MalwareNet including the number of neurons in each layer and the fully connected layer size. With MalwareNet obtaining especially remarkable results, the examination of our models shows great accuracy, precision, recall, and F1 scores. We examine the training accuracy, F1 score, and log loss measures in great detail; we then show the classification performance over all classes using confusion matrices. With notable increases in malware detection and categorization, the suggested model offers a strong and practical answer for uses in cybersecurity. Supported by diligent hyperparameter adjustment and performance evaluation, this work expands the field by providing a comprehensive comparison of many deep learning architectures and how effectively they classify malware. The results show how well deep learning models might improve malware detection systems and protect digital environments against hostile attacks.

关键词： Deep learning Training Performance evaluation Accuracy Neurons Malware Loss measurement Computer security Kernel Optimization

来源：评论

学校读者我要写书评

暂无评论

Two-stream Multi-level Dynamic Point Transformer for Two-person Interaction Recognition

arXiv

引用

arXiv 2023年

作者： Liu, Yao Cui, Gangfeng Luo, Jiahui Chang, Xiaojun Yao, Lina School of Computer Science and Engineering University of New South Wales SydneyNSW2052 Australia School of Computing and Information Systems University of Melbourne MelbourneVIC3010 Australia Faculty of Engineering and Information Technology University of Technology Sydney SydneyNSW2007 Australia Data 61 CSIRO School of Computer Science and Engineering University of New South Wales SydneyNSW2015 Australia

As a fundamental aspect of human life, two-person interactions contain meaningful information about people's activities, relationships, and social settings. Human action recognition serves as the foundation for many smart applications, with a strong focus on personal privacy. However, recognizing two-person interactions poses more challenges due to increased body occlusion and overlap compared to single-person actions. In this paper, we propose a point cloud-based network named Two-stream Multi-level Dynamic Point Transformer for two-person interaction recognition. Our model addresses the challenge of recognizing two-person interactions by incorporating local-region spatial information, appearance information, and motion information. To achieve this, we introduce a designed frame selection method named Interval Frame Sampling (IFS), which efficiently samples frames from videos, capturing more discriminative information in a relatively short processing time. Subsequently, a frame features learning module and a two-stream multi-level feature aggregation module extract global and partial features from the sampled frames, effectively representing the local-region spatial information, appearance information, and motion information related to the interactions. Finally, we apply a transformer to perform self-attention on the learned features for the final classification. Extensive experiments are conducted on two large-scale datasets, the interaction subsets of NTU RGB+D 60 and NTU RGB+D 120. The results show that our network outperforms state-of-the-art approaches in most standard evaluation settings. Copyright © 2023, The Authors. All rights reserved.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Computational access point selection based on resource allocation optimization to reduce the edge computing latency

引用

Measurement: Sensors 2022年 24卷

作者： Kumaran, K. Sasikala, E. Research Scholar SRM Institute of Science and Technology Chennai600089 India Department of Information Technology Easwari Engineering College Chennai600089 India Department of Data Science & Business Systems SRM Institute of Science and Technology Kattankulathur Chennai603203 India

Mobile Devices (MD) and the Internet of Things (IoT) transfer compute tasks to the client through Computational Access Point (CAP) through the extensive implementation of Local Area Networks (LAN). If all IoT or mobile devices choose the same ports of entry to offload their work, then the calculation of offloading leads to an increase in operating costs. In a network of different CAP to Edge Computing (EC), this research proposes a resource allocation optimization technique and a computation offloading strategy that aims to reduce network costs by recommending the best transmit power distribution, bandwidth assignment, and computation task planning. The optimization problem of NP-hard solved by the proposed technique separates from sub-problems of resource allocation and offloading technique. An algorithm called Dynamic Weighed Quantum Arithmetic Optimization Algorithm (DWQAOA) was proposed for this offloading problem NP-hard. To achieve the ideal solution, a transition probability introduced the optimal Pareto connection during the optimization process. Numerous simulation results show that the edge task latency could be reduced from 3 to 25% relative to competing solutions. © 2022 The Authors

关键词： Resource allocation

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：