检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Wen, Jinbo Kang, Jiawen Zhang, Yang Zhong, Yue Niyato, Dusit Xu, Jie Tang, Jianhang Yuen, Chau College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics China School of Automation Guangdong University of Technology China College of Computing and Data Science Nanyang Technological University Singapore Guangdong Provincial Key Laboratory of Future Networks of Intelligence The Chinese University of Hong Kong Shenzhen China State Key Laboratory of Public Big Data Guizhou University China School of Electrical and Electronics Engineering Nanyang Technological University Singapore

Mobile metaverses have attracted significant attention from both academia and industry, which are envisioned as the next-generation Internet, providing users with immersive and ubiquitous metaverse services through mobile devices. Driven by Large Language Models (LLMs) and Vision-Language Models (VLMs), Artificial Intelligence (AI) agents hold the potential to empower the creation, maintenance, and evolution of mobile metaverses. Currently, AI agents are primarily constructed using cloud-based LLMs and VLMs. However, several challenges hinder their effective implementation, including high service latency and potential sensitive data leakage during perception and processing. In this paper, we develop an edge-cloud collaboration-based federated AI agent construction framework in mobile metaverses. Specifically, Edge Servers (ESs), acting as agent infrastructures, collaboratively create agent modules in a distributed manner. The cloud server then integrates these modules into AI agents and deploys them at the edge, thereby enabling low-latency AI agent services for users. Considering that ESs may exhibit dynamic levels of willingness to participate in federated AI agent construction, we design a two-period dynamic contract model to continuously motivate ESs to participate in agent module creation, effectively addressing the dynamic information asymmetry between the cloud server and the ESs. Furthermore, we propose an Enhanced Diffusion Model-based Soft Actor-Critic (EDMSAC) algorithm to efficiently generate optimal dynamic contracts, in which dynamic structured pruning is applied to DM-based actor networks to enhance denoising efficiency and policy learning performance. Extensive simulations demonstrate the effectiveness and superiority of the EDMSAC algorithm and the proposed contract model. Copyright © 2025, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Online Container Scheduling for Low-Latency IoT Services in Edge Cluster Upgrade: A Reinforcement Learning Approach

Online Container Scheduling for Low-Latency IoT Services in ...

引用

2023 IEEE/CIC International Conference on Communications in China, ICCC 2023

作者： Cui, Hanshuai Tang, Zhiqing Lou, Jiong Jia, Weijia Beijing Normal University Institute of Artificial Intelligence and Future Networks Zhuhai519087 China Beijing Normal University School of Artificial Intelligence Beijing100875 China Shanghai Jiao Tong University Department of Computer Science and Engineering Shanghai200240 China BNU-HKBU United International College Guangdong Key Lab of Ai and Multi-Modal Data Processing Zhuhai519087 China

ISBN: (纸本)9798350345384

In Mobile Edge Computing (MEC), Internet of Things (IoT) devices offload computationally-intensive tasks to edge nodes, where they are executed within containers, reducing the reliance on centralized cloud infrastructure. Frequent up-grades are essential to maintain the efficient and secure operation of edge clusters. However, traditional cloud cluster upgrade strategies are ill-suited for edge clusters due to their geographically distributed nature and resource limitations. Therefore, it is crucial to properly schedule containers and upgrade edge clusters to minimize the impact on running tasks. In this paper, we propose a low-latency container scheduling algorithm for edge cluster upgrades. Specifically: 1) We formulate the online container scheduling problem for edge cluster upgrade to minimize the total task latency. 2) We propose a policy gradient-based reinforcement learning algorithm to address this problem, considering the unique characteristics of MEC. 3) Experimental results demonstrate that our algorithm reduces total task latency by approximately 27% compared to baseline algorithms. © 2023 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Crowdsourcing for data Labeling with Heterogeneous Crowd

Hierarchical Crowdsourcing for Data Labeling with Heterogene...

引用

International Conference on data engineering

作者： Haodi Zhang Wenxi Huang Zhenhan Su Junyang Chen Di Jiang Lixin Fan Chen Zhang Defu Lian Kaishun Wu College of Computer Science and Software Engineering Shenzhen University Department of Computing Hong Kong Polytechnic University WeBank Institute of Financial Technology Shenzhen University Webank Co. Ltd. School of Data Science University of Science and Technology of China

With the rapid and continuous development of data-driven technologies such as supervised learning, high-quality labeled data sets are commonly required by many applications. Due to the easiness of crowdsourcing small tasks with low cost, a straightforward solution for label quality improvement is to collect multiple labels from a crowd, and then aggregate the answers. The aggregation strategies include majority voting and its many variants, EM-based approaches, Graph Neural Nets and so on. However, due to the uncertainty information loss and commonly existing task correlations, the aggregated labels usually contain errors and may damnify the downstream model *** address the above problem, we propose a hierarchical crowdsourcing framework 1 for data labeling with noisy answers about correlated data. We make use of the heterogeneity of the labeling crowd and form an initialization-checking-update loop to improve the quality of labeled data. We formalize and successfully solve the core optimization problem, namely, selecting a proper set of checking tasks for each round. We prove that maximizing the expected quality improvement is equivalent to minimizing the conditional entropy of the observations given the crowdsourced answer families for the selected task set, which is NP-hard to solve. Therefore, we design an efficient approximation algorithm and conduct a series of experiments on real data. The experimental results show that the proposed method effectively improves the quality of the labeled data sets as well as the SOTA performance, yet without extra human labor costs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Transformer Based Sales Prediction of Smart Container in New Retail Era 21

A Transformer Based Sales Prediction of Smart Container in N...

引用

5th International Conference on Deep Learning Technologies, ICDLT 2021

作者： Jin, Ying Gao, Ming Yu, Jixiang School of Management Science and Engineering Dongbei University of Finance and Economics Liaoning Dalian China Center for Post-doctoral Studies of Computer Science Northeastern University Shenyang China School of Data Science and Artificial Intelligence Dongbei University of Finance and Economics Liaoning Dalian China

ISBN: (纸本)9781450390163

With the advent of the new retail era, the value of unmanned smart container is increasingly prominent. Fast and flexible self-service is favored by consumers. How to use accumulated historical sales data to predict sales in the future is an important part of smart container operation management. Reasonable sales prediction can not only reduce the inventory cost, but also reduce the shortage loss of the container. Based on the smart container sales data of Dalian Xiaode New Retail Co., Ltd., through detailed exploratory analysis in many aspects, this paper carries out the feature selection of sales prediction, and uses random forest, XGBoost, Transformer and other algorithms to predict sales. The experimental results show that the prediction accuracy of Transformer is better than traditional algorithms, whose MAPE is 14.67% lower than that of the worst one. Transformer can be well applied in the field of sales prediction of smart container. And in this experiment, compared with Transformer using sine and cosine functions for positional encoding, Transformer encoded by position index has better prediction performance and stronger stability. © 2021 ACM.

关键词： Containers

来源：评论

学校读者我要写书评

暂无评论

Optimization for Task Offloading and Downloading in UAV-Assisted MEC Systems with Aerial to Aerial Collaboration

Optimization for Task Offloading and Downloading in UAV-Assi...

引用

International Conference on computer Supported Cooperative Work in Design

作者： Jiawei Jiang Xiang Tian Yubing Han Chunyu Hu Bin Feng Jiguo Yu Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China School of Information Science and Technology Taishan University Taian China School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China Big Data Institute Qilu University of Technology Jinan China

ISBN: (数字)9798331513054

ISBN: (纸本)9798331513061

Owing to the easy deployment and mobile flexibility, Unmanned Aerial Vehicle (UAV) assisted Mobile Edge Computing (MEC) has been deemed as one potential technology for handling the computation-intensive tasks at terminal devices (TDs). In this work, a MEC architecture assisted by UAVs is designed which achieves efficient offloading, computing, and downloading for tasks from multiple TDs via aerial to aerial collaboration of two UAVs. In this architecture, the task offloading process contains two parts, i.e., the offloading from TDs to a mobile UAV which flies around TDs, and the offloading from the mobile UAV to a hovering UAV which hovers in the air. The computing tasks from TDs will be divided into three parts allocated to the TDs themselves, and both two UAVs. Upon completion of computation, the computation results are downloaded to the TDs. The optimization objective is to seek for an optimal task division strategy to attain the weighted total energy consumption minimization for all devices. Since the formulated optimization problem is not convex, we develop a two-step iteration algorithm which jointly optimizes computing frequency, task allocation volume, as well as UAV's trajectory based on the method of block coordinate descent. Simulation results confirm the effectiveness and performance advantages of the designed algorithm.

关键词： Energy consumption Simulation Collaboration computer architecture Autonomous aerial vehicles Minimization Resource management Jamming Trajectory optimization Radio spectrum management

来源：评论

学校读者我要写书评

暂无评论

Health Information Broadcast Distributed Pattern Association based on Estimated Volume

Health Information Broadcast Distributed Pattern Association...

引用

Artificial Intelligence and Smart Systems (ICAIS), International Conference on

作者： R. Dhanalakshmi A.V. Kalpana J. Umamageswaran B. Praveen Kumar Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai Tamil Nadu India Department of Data Science and Business Systems SRM Institute of Science and Technology Kattankulathur Chennai Tamil Nadu India Department of Information Technology R.M.K Engineering College Chennai Tamil Nadu India Department of Computer Science and Engineering K L University Andhra Pradesh India

Any study on dataset and inadequate representation are featured in the conventional data mining process. Consequently, data mining findings reveal high chances of challenges and huge root-mean-square approximation errors besides long utilization. The proposed algorithm for associating distribution patterns in streams of medical data is based on density calculations. The distance approach will be utilized to identify any orphaned or questionable data within the data stream, and the gathering of medical data will be used to achieve the aim. data stream duplicates are compiled by sector match quality, and both the out-of-the-ordinary and the redundant data are eventually eliminated. The density of the information stream is then calculated based on histogram calculations. This investigates the flow of medical information from the viewpoints of concentration and dispersion, as well as the physical characteristics of information distribution, in combination with the results of data density estimate. The model is then built upon a neural network composite, the data distribution parameters in the clustering layer of the model are entered, and in-depth testing is carried out via the neural B.P. network (Back Propagation) on the mining layer of the model. Temporarily, all laws under the grouping of the hidden layer operation value besides the resulting output value are derived, and all rules under the grouping of the hidden layer value besides the conforming contribution value in the medical information stream pattern. The new findings show a curve of contour contiguous to the true curve of the likelihood density, a fair range of medical data dispersion and high medical data consistency, a lower probability of data redundancy, a low RMSEA for mining, a lower time for data recovery.

关键词： Histograms Neural networks Redundancy Approximation algorithms data models Classification algorithms data mining

来源：评论

学校读者我要写书评

暂无评论

data-Driven Computational Algorithms for Predicting Electricity Consumption Missing Values: A Comparative Study

Data-Driven Computational Algorithms for Predicting Electric...

引用

Australasian Universities Power engineering Conference, AUPEC

作者： Bavly Hanna Guandong Xu Xianzhi Wang Jahangir Hossain Department of Computer Science University of Technology Sydney Sydney Australia School of Electrical and Data Engineering University of Technology Sydney Sydney Australia

data availability is a significant issue and barrier for modeling and analyzing low voltage networks. This paper develops, implements, and compares several prediction algorithms for finding missing values in energy usage for commercial consumers. Four predictive machine learning models, such as random forest regression, linear regression, multi-layer perceptron, and decision trees, are utilized in this paper. Four commercial users from a regional city in Australia are selected as a dataset based on 30-minute intervals. Firstly, the obtained data is analyzed and pre-processed and then utilized for model training and testing. RMSE and MAE measures are used to compare the effectiveness of each machine learning model. This paper concludes that the multi-layer perceptron model provides better performance than that of random forest, decision tree, and linear regression. The RMSE and MAE of MLP model are 4.7472 and 4.2103, respectively, when using individual users as a training set.

关键词：

来源：评论

学校读者我要写书评

暂无评论

AI-Driven Energy Optimization: Household Power Consumption Prediction With LSTM Networks and PyTorch-Ray Tune in Smart IoT Systems

AI-Driven Energy Optimization: Household Power Consumption P...

引用

International Conference on Microelectronics, ICM

作者： Vinoth Kumar Kolluru Yagnesh Challagundla Advaitha Naidu Chintakunta Bappadittya Roy Amine Bermak Renuka Devi S M Department of Data Science Stevens Institute of Technology Hoboken NJ USA Engineering Education Dept. University of Florida USA Department of Computer science and Engineering University of North Carolina Charlotte USA School of Electronics Engineering VIT-AP University Amaravati Andhra Pradesh India College of Science and Engineering Hamad bin Khalifa Universityy Ar-Rayyan Qatar Dept. of ECE G Narayanamma Institute of Technology & Science (for women) Hyderabad

ISBN: (数字)9798350379396

ISBN: (纸本)9798350379402

Energy optimization is critical in smart home systems and IoT networks, necessitating innovative models that reduce energy use. This research presents two Long Short-Term Memory (LSTM) network models for estimating home power consumption using time-series data. The first model employs a standard LSTM technique with multivariate characteristics including global active power, voltage, and sub-metering variables. This technique strikes a balance between prediction accuracy and computing economy, making it ideal for resource-constrained applications. The second model provides an optimized version of the LSTM, which uses PyTorch and Ray Tune for hyperparameter optimization. The optimization focuses on tweaking learning rates, batch sizes, and LSTM layers to improve model accuracy and convergence speed. Using hyperparameter tuning, the Mean Squared Error (MSE) for global active power forecasts is decreased to 0.0018, proving its appropriateness for low-resource IoT systems. Both models are tested on a resampled real-world household electric power consumption dataset for effective training. The study emphasizes the advantages of multivariate time-series analysis and hyperparameter optimization, demonstrating that the optimized LSTM model can accurately predict energy consumption, enhance energy management in smart homes, and lower computing costs. Future work will look into combining more IoT data streams and real-world deployment for further improvement. The findings add to the expanding body of knowledge about energy optimization in IoT environments, addressing the crucial demand for effective, real-time energy management systems.

关键词： Power demand Accuracy Computational modeling Biological system modeling Smart homes Predictive models Hyperparameter optimization Real-time systems Long short term memory Optimization

来源：评论

学校读者我要写书评

暂无评论

NAGNet: A Convolutional Neural Network for Real-Time Sentiment Analysis of Students

NAGNet: A Convolutional Neural Network for Real-Time Sentime...

引用

IEEE International Symposium on Information (IT) in Medicine and Education, ITME

作者： Huihui Zhu Pengyun Hu Xianpiao Tang Daoxun Xia School of Big Data and Computer Science Guizhou Normal University Guiyang China Engineering Laboratory for Applied Technology of Big Data in Education Guizhou Normal University Guiyang China

Students' emotional state during class can reflect their understanding of classroom knowledge. Analysing the emotional data of students is helpful for ensuring that teachers understand students' learning effect and improve their teaching quality. With the rapid development of information education, student emotion analysis has been widely studied in the field of educational technology. However, due to the complexity and inconsistency of educational scenes, traditional facial expression recognition algorithms have low accuracy, losing information while extracting facial expression features, and student emotion analysis faces several technical issues. Therefore, based on the theory of deep learning, this paper proposes a novel network model NAGNet for recognizing and analysing students' expressions in classroom environments, to assist teachers in carrying out reasonable teaching effect evaluations. In this model, Res2Net50 is used as the backbone network to enhance the feature extraction ability of the model. Moreover, non-local attention module is added to integrate the global facial expression feature information to achieve fine-grained sentiment analysis. In addition, we replace the original pooling algorithm with generalized mean (GeM) pooling to pool the features of the last layer. The experimental results on the public affection dataset FERPlus, which includes eight emotions, show that the NAGNet model achieves an accuracy of 89.3%. Thus, the proposed model has higher accuracy than existing methods. In a real-world application scenario, the NAGNet model can analyse and detect students' emotions in real time. What's more, it has good robustness.

关键词： Deep learning Training Analytical models Sentiment analysis Emotion recognition Face recognition Feature extraction

来源：评论

学校读者我要写书评

暂无评论

LEARNING EXPENSIVE COORDINATION: AN EVENT-BASED DEEP RL APPROACH 8

LEARNING EXPENSIVE COORDINATION: AN EVENT-BASED DEEP RL APPR...

引用

8th International Conference on Learning Representations, ICLR 2020

作者： Yu, Runsheng Wang, Xinrun Wang, Rundong Zhang, Youzhi An, Bo Shi, ZhenYu Lai, Hanjiang School of Computer Science and Engineering Nanyang Technological University Singapore School of Data and Computer Science Sun Yat-sen University Guangzhou China

Existing works in deep Multi-Agent Reinforcement Learning (MARL) mainly focus on coordinating cooperative agents to complete certain tasks jointly. However, in many cases of the real world, agents are self-interested such as employees in a company and clubs in a league. Therefore, the leader, i.e., the manager of the company or the league, needs to provide bonuses to followers for efficient coordination, which we call expensive coordination. The main difficulties of expensive coordination are that i) the leader has to consider the long-term effect and predict the followers' behaviors when assigning bonuses, and ii) the complex interactions between followers make the training process hard to converge, especially when the leader's policy changes with time. In this work, we address this problem through an event-based deep RL approach. Our main contributions are threefold. (1) We model the leader's decision-making process as a semi-Markov Decision Process and propose a novel multi-agent event-based policy gradient to learn the leader's long-term policy. (2) We exploit the leader-follower consistency scheme to design a follower-aware module and a follower-specific attention module to predict the followers' behaviors and make accurate response to their behaviors. (3) We propose an action abstraction-based policy gradient algorithm to reduce the followers' decision space and thus accelerate the training process of followers. Experiments in resource collections, navigation, and the predator-prey game reveal that our approach outperforms the state-of-the-art methods dramatically. © 2020 8th International Conference on Learning Representations, ICLR 2020. All rights reserved.

关键词： Reinforcement learning

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：