检索结果-内蒙古大学图书馆

DEEPCON: Improving Distributed Deep Learning Model Consistency in Edge-Cloud Environments via Distillation

IEEE Transactions on Cognitive Communications and Networking 2025年

作者： Qian, Bin Qian, Jiaxu Wen, Zhenyu Wu, Di He, Shibo Chen, Jiming Ranjan, Rajiv Zhejiang University State Key Laboratory of Industrial Control Technology Hangzhou China Zhejiang University of Technology Institute of Cyberspace Security College of Information Engineering Hangzhou China University of St Andrews School of Computer Science United Kingdom Newcastle University United Kingdom

In a typical distributed Deep Learning (DL) based application, models are configured differently to meet the requirements of resource constraints. For instance, a large ResNet56 model is deployed on the cloud server while a small lightweight MobileNet model is more suitable for the end-user device with fewer computation resources. However, the heterogeneity of the model architectures and configurations may bring a systemic problem -models may produce different outputs when given the same input. This inconsistency problem may cause severe system failure of prediction agreement inside the application. Current research has not studied the systemic design for efficiently detecting and reducing the inconsistency among models in distributed DL applications. With the increasing scale of distributed DL applications, the challenges of inconsistency mitigation should consider both algorithm and system design. To this end, we design and implement DEEPCON, an adaptive deployment system across the edge-cloud layer with over-the-air model updates. We implement ASRS sampling for efficiently sampling data to reveal the real data distribution as well as model prediction inconsistency. Then, we implement DMML-Par, an asynchronous parallel training algorithm for quickly updating the models and reducing inconsistency. implements over-the-air updates with a set of APIS to enable seamless inconsistency detection and reduction in such deep learning applications. Our experiment results on both vision and language tasks demonstrate that DMML could improve the model consistency up to 4%, 7%, and 13% at CIFAR10/100 and IMDB datasets without sacrificing the accuracy of individual models. We also show that the ASRS sampling can save 90% network bandwidth of data transmission and that DMML-Par is up to 60% faster compared to simple synchronous parallel training. 2015 IEEE.

关键词： Distributed cloud

来源：评论

学校读者我要写书评

暂无评论

Conservative Q-Learning Based Furnace Temperature control in Municipal Solid Waste Incineration Process

Conservative Q-Learning Based Furnace Temperature Control in...

引用

International Conference on Industrial Artificial Intelligence (IAI)

作者： Hao Tian Jian Tang Heng Xia Jian Zhang Tianzheng Wang Wen Yu Faculty of Information Technology Beijing University of Technology Beijing China School of Computer Science Nanjing University of Information Technology Nanjing China Departamento de Control Automatico CINVESTAV-IPN (National Polytechnic Institute) Mexico City Mexico

ISBN: (数字)9798350356618

ISBN: (纸本)9798350356625

The nonlinear, time-varying, and lagging nature of the municipal solid waste incineration (MSWI) process presents challenges for ensuring controller safety. While offline reinforcement learning (RL) can ensure safety in furnace temperature (FT) control, its performance is hindered by extrapolation errors, making it unsuitable for direct application in the incineration environment. To address this, we propose a conservative Q-learning-based furnace temperature control strategy (CQL-FTC). This strategy involves two stages: online sampling and offline training. During the online sampling stage, the agent interacts with the environment to collect samples, building an experience replay buffer (ERB) and performing pretraining. In the offline training stage, we introduce the CQL method, adding constraint terms to the traditional Bellman equation to minimize extrapolation errors. After offline training, the agent is directly applied to the FT setpoint control in the incineration process. Simulation results using the actual MSWI process dataset demonstrate the effectiveness of the proposed method in complex industrial environments.

关键词： Training Waste management Waste materials Extrapolation Temperature distribution Incineration Furnaces Process control Temperature control Safety

来源：评论

学校读者我要写书评

暂无评论

Adjusted and controllability Properties of Memristor Rucklidge System

Adjusted and Controllability Properties of Memristor Rucklid...

引用

2023 IEEE International Conference on Memristive Computing and Applications, ICMCA 2023

作者： Zang, Hongyan Huang, Lili Zhang, Shourong Joseph, Annie Anak Fu, Haiyan Lei, Tengfei Qilu Institute of Technology Collaborative Innovation Center of Memristive Computing Application Jinan250200 China Universiti Malaysia Sarawak Faculty of Engineering Kota Samarahan Sarawak Malaysia College of Intelligent Manufacturing and Control Engineering Qilu Institute of Technology Jinan250200 China Shandong Vanform High Energy Physics Technology Co. Ltd. Jinan250100 China Universiti Malaysia Sarawak Department of Electrical and Electronics Kota Samarahan Sarawak Malaysia College of Computer and Information Engineering Qilu Institute of Technology Jinan250200 China Jinan250200 China

ISBN: (纸本)9798350316230

There are many models of memristors, but there are no reports on exponential memristors. In this paper, a three- dimension memristor Rucklidge system with exponential memristor is proposed by introducing exponential function and constructing exponential memristor. The dynamic behavior of the system is analyzed by means of Lyapunov exponent and bifurcation diagram. The complex attractor coexistence and amplitude modulation characteristics of the system are found. The internal variable of memristor is introduced into the second two-dimensional equation as a feedback term, and a conditional symmetric system is constructed for the three-dimensional memristor system, which can realize the polarity control of output signal. By discussing the dynamic characteristics of exponential memristor chaotic system and the control mechanism of amplitude modulation, bias and conditional symmetry, the complex dynamic characteristics of exponential function memristor are revealed, which lays a foundation for its application in the field of secure communication. © 2023 IEEE.

关键词： Memristors

来源：评论

学校读者我要写书评

暂无评论

MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at Scale 39

MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at ...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Andreychuk, Anton Yakovlev, Konstantin Panov, Aleksandr Skrynnik, Alexey AIRI Moscow Russian Federation Federal Research Center “Computer Science and Control” The Russian Academy of Sciences Moscow Russian Federation Moscow Institute of Physics and Technology Dolgoprudny Russian Federation

ISBN: (纸本)157735897X

Multi-agent pathfinding (MAPF) is a problem that generally requires finding collision-free paths for multiple agents in a shared environment. Solving MAPF optimally, even under restrictive assumptions, is NP-hard, yet efficient solutions for this problem are critical for numerous applications, such as automated warehouses and transportation systems. Recently, learning-based approaches to MAPF have gained attention, particularly those leveraging deep reinforcement learning. Typically, such learning-based MAPF solvers are augmented with additional components like single-agent planning or communication. Orthogonally, in this work we rely solely on imitation learning that leverages a large dataset of expert MAPF solutions and transformer-based neural network to create a foundation model for MAPF called MAPF-GPT. The latter is capable of generating actions without additional heuristics or communication. MAPF-GPT demonstrates zero-shot learning abilities when solving the MAPF problems that are not present in the training dataset. We show that MAPF-GPT is able to outperform the state-of-the-art learnable MAPF solvers on a diverse range of problem instances and is computationally efficient during inference. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Bias Detection and Prediction of Mapping Errors in Camera Calibration 1

引用

42nd German Conference on Pattern Recognition, DAGM GCPR 2020 held in parallel with 25th International Symposium on Vision, Modeling, and Visualization, VMV 2020 and 10th Eurographics Workshop on Visual Computing for Biology and Medicine, VCBM 2020

作者： Hagemann, Annika Knorr, Moritz Janssen, Holger Stiller, Christoph Robert Bosch GmbH Corporate Research Computer Vision Research Lab Hildesheim Germany Institute of Measurement and Control Karlsruhe Institute of Technology Karlsruhe Germany

ISBN: (数字)9783030712785

ISBN: (纸本)9783030712778

Camera calibration is a prerequisite for many computer vision applications. While a good calibration can turn a camera into a measurement device, it can also deteriorate a system’s performance if not done correctly. In the recent past, there have been great efforts to simplify the calibration process. Yet, inspection and evaluation of calibration results typically still requires expert knowledge. In this work, we introduce two novel methods to capture the fundamental error sources in camera calibration: systematic errors (biases) and remaining uncertainty (variance). Importantly, the proposed methods do not require capturing additional images and are independent of the camera model. We evaluate the methods on simulated and real data and demonstrate how a state-of-the-art system for guided calibration can be improved. In combination, the methods allow novice users to perform camera calibration and verify both the accuracy and precision. © 2021, Springer Nature Switzerland AG.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Prediction of Heat Transfer in Smart Buildings using Internet of Things based Artificial Intelligence 3

Prediction of Heat Transfer in Smart Buildings using Interne...

引用

3rd International Conference on Electronics and Sustainable Communication Systems, ICESC 2022

作者： Rehman, TasneemBano Madiajagan, M. Ramya, D. Saratha, M. Mohan Krishna, P.B. Kumar, Prashant Sage University Department of Computer Science and Engineering Madhya Pradesh Bhopal India Vellore Institute of Technology School of Computer Science and Engineering Tamil Nadu Vellore India Sathyabama Institute of Science and Technology Department of Eee Tamil Nadu Chennai India M.Kumarasamy College of Engineering Department of Mca Tamil Nadu Karur India National Institute of Technology Department of Process Control Telangana Warangal India Lovely Professional University Department of Chemical Engineering Punjab India

ISBN: (数字)9781665479714

ISBN: (纸本)9781665479714

With rapid growth and a higher standard of living, the demand for usable energy has increased tremendously over the last few decades, with the construction industry being one of the most notable examples. The energy efficiency of a building is influenced by a variety of factors, including the weather in the area, the structure characteristics, and how people use electricity. These algorithms can be used to generate a prediction model for smart building indoor temperature forecasting using the Internet of Things and several commonly used artificial intelligence algorithms, which can then be used to generate a prediction model for smart building outdoor temperature forecasting. It was necessary to employ freely accessible software tools in order to make geometric and sensory data available in a virtual reality application and to build the PMV and PPD indices. On a low-cost edge computing device, semi-automatic analysis of streaming thermal images from a non-intrusive sensor was performed in order to offer near real-time Mean Radiant Temperature (MRT)estimations. An experimental space that had been mechanically preconditioned was filled with a dynamic temperature environment, and a prototype of the system was employed in a series of trials. Results show that the data backs up the fact that the system output and the heat sensations felt under different settings are consistent. © 2022 IEEE.

关键词： Application programs

来源：评论

学校读者我要写书评

暂无评论

Advancing Medical Innovation through Blockchain-Secured Federated Learning for Smart Health

引用

IEEE Journal of Biomedical and Health Informatics 2025年 PP卷 PP页

作者： Khan, Salabat Khan, Mansoor Khan, Muhammad Asghar Wang, Lu Wu, Kaishun Qilu Institute of Technology School of Computer and Information Engineering Jinan China Qilu Institute of Technology School of Intelligent Manufacturing and Control Engineering Jinan China Prince Mohammad Bin Fahd University Department of Electrical Engineering P.O. Box 1664 Al-Khobar31952 Saudi Arabia Shenzhen University College of Computer Science and Software Engineering Shenzhen China Thrust Guangzhou China

The rapid digitization of healthcare systems has led to a vast accumulation of electronic medical records (EMRs), offering an invaluable source of patient data that can significantly advance medical research and improve patient care. However, sharing EMRs for research purposes presents challenges, particularly concerning data privacy, security, and the limitations of traditional centralized data-sharing models. This paper introduces a novel approach that leverages blockchain technology to facilitate federated learning with EMRs, thereby addressing these challenges. Federated learning enables multiple institutions to collaboratively train a robust machine learning model without sharing raw data, preserving privacy and security. By integrating blockchain, this framework enhances data integrity, immutability, and trust, all in a decentralized environment. The blockchain serves as a transparent and secure ledger, recording model updates and aggregating them through a consensus-based mechanism. Smart contracts further enforce data usage policies, allowing only authorized access and maintaining control over data ownership and sharing. This approach empowers medical researchers and institutions to collaborate more effectively, accelerating the discovery of treatments, advancements in personalized medicine, and insights into rare diseases. It also enables patients to contribute to medical research while retaining control over their personal data, fostering a patient-centered approach to healthcare innovation. Experimental results confirm the efficacy and efficiency of this blockchain-enabled federated learning framework, highlighting its potential to transform medical research and adhere to stringent privacy and security standards. This study emphasizes the pivotal role of blockchain in enhancing big data analytics within healthcare, paving the way for improved collaboration, innovation, and patient outcomes. © 2025 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Rethinking Graph Transformer Architecture Design for Node Classification

arXiv

引用

arXiv 2024年

作者： Zhou, Jiajun Chen, Xuanze Xie, Chenxuan Yu, Shanqing Xuan, Qi Yang, Xiaoniu Institute of Cyberspace Security Zhejiang University of Technology Hangzhou310023 China Binjiang Institute of Artificial Intelligence ZJUT Hangzhou310056 China College of Computer Science and Technology Zhejiang University of Technology Hangzhou310023 China Science and Technology on Communication Information Security Control Laboratory Jiaxing314033 China

Graph Transformer (GT), as a special type of Graph Neural Networks (GNNs), utilizes multi-head attention to facilitate high-order message passing. However, this also imposes several limitations in node classification applications: 1) nodes are susceptible to global noise;2) self-attention computation cannot scale well to large graphs. In this work, we conduct extensive observational experiments to explore the adaptability of the GT architecture in node classification tasks and draw several conclusions: the current multi-head self-attention module in GT can be completely replaceable, while the feed-forward neural network module proves to be valuable. Based on this, we decouple the propagation (P) and transformation (T) of GNNs and explore a powerful GT architecture, named GNNFormer, which is based on the P/T combination message passing and adapted for node classification in both homophilous and heterophilous scenarios. Extensive experiments on 12 benchmark datasets demonstrate that our proposed GT architecture can effectively adapt to node classification tasks without being affected by global noise and computational efficiency limitations. Copyright © 2024, The Authors. All rights reserved.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Distributed Nash Equilibrium Seeking with Stochastic Event-Triggered Mechanism

arXiv

引用

arXiv 2023年

作者： Huo, Wei Tsang, Kam Fai Elvis Yan, Yamin 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 KTH Royal Institute of Technology Stockholm Sweden

In this paper, we study the problem of consensus-based distributed Nash equilibrium (NE) seeking where a network of players, abstracted as a directed graph, aim to minimize their own local cost functions non-cooperatively. Considering the limited energy of players and constrained bandwidths, we propose a stochastic event-triggered algorithm by triggering each player with a probability depending on certain events, which improves communication efficiency by avoiding continuous communication. We show that the distributed algorithm with the developed event-triggered communication scheme converges to the exact NE exponentially if the underlying communication graph is strongly connected. Moreover, we prove that our proposed event-triggered algorithm is free of Zeno behavior. Finally, numerical simulations for a spectrum access game are provided to illustrate the effectiveness of the proposed mechanism by comparing it with some existing event-triggered methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

A Vein Segmentation Method Based on Improved U-Net 12

A Vein Segmentation Method Based on Improved U-Net

引用

12th International Conference of Information and Communication technology, ICTech 2023

作者： Wang, Hongliang Zhu, Liang Deng, Changyi Xiu, Pengfei Shenyang Institute of Computing Technology Chinese Academy of Sciences Shenyang China University of Chinese Academy of Sciences Beijing China Liaoning Province Human-Computer Interaction System Engineering Research Center Based on Digital Twin Shenyang China Institute of Software China Industrial Control Systems Cyber Emergency Response Team Beijing100040 China

ISBN: (纸本)9798350332919

Recently, medical research has revealed that diffusion weighted imaging (DWI) is less sensitive than susceptibility-weighted imaging (SWI) for acute ischemic stroke. Brain vein analysis in SWI is very important for personalised therapy, quantitative diagnosis, and prognosis of acute ischemic stroke. Therefore, the accurate segmentation of veins in SWI images has important significance in supplementary diagnosis and medical research. However, vein segmentation in SWI is a challenging task because different veins show distinct variability and uneven intensity and are difficult to manually split. This paper proposes an improvement method based on U-net, by using dense connection and mixing loss function, the best Dice coefficient and the ASSD were obtained by other comparison methods. The method presented in this article can be used as a potential tool to accurately measure SWI cerebral veins and help physicians in decision making. © 2023 IEEE.

关键词： Image segmentation

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：