检索结果-内蒙古大学图书馆

Large Model Collaborative Optimization Enhanced by Mobile Consumer Electronics Constructed Federated ML

IEEE Transactions on Consumer Electronics 2024年

作者： Bu, Chao Li, Jianlong Wang, Jinsong Tianjin University of Technology Key Laboratory of Computer Vision and System of Ministry of Education School of Computer Science and Engineering Tianjin300384 China Tianjin University of Technology Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology School of Computer Science and Engineering Tianjin300384 China

With the rapid development of internet of everything, current Mobile Consumer Electronics (MCEs) can support complex computing due to their specialized data-handling capacity. Thus, federated Machine Learning (ML) and edge computing are integrated to leverage such capacity of MCEs to enhance large model training. This paper studies motivations of all participants that take part in large model collaborative optimization from a global perspective, and proposes the positive promotion pattern that collaborates MCEs, Edge Computing Servers (ECSs), and the centralized computing center to support large model collaborative optimization by introducing the idea of computing economic. We firstly propose the novel three-layered framework for large model collaboratively optimizing based on MCEs constructed federated ML. Secondly, we introduce the economic factor into the large model optimization by devising algorithms to assess consumptions of all participants and dynamically adjust incentive benefits according to the qualities of trained models. Finally, we design the scheme to encourage more MCEs to train local models by adaptively adjusting the ECS pricing strategy with the ECS future economic benefit trend considered. Experimental results validate that our proposed approach is able to optimize the global accuracy and improve the economic benefit more efficiently than the state of the art. © 1975-2011 IEEE.

关键词： Mobile edge computing

来源：评论

学校读者我要写书评

暂无评论

Using Lock-Free Design for Throughput-Optimized Cache Eviction

引用

Proceedings of the ACM on Measurement and Analysis of Computing systems 2025年第2期9卷 1-28页

作者： Dong, Chao Wang, Fang Jiang, Hong Feng, Dan Huazhong University of Science and Technology Key Laboratory of Information Storage System Wuhan National Laboratory for Optoelectronics Engineering Research Center of data storage systems and Technology School of Computer Science and Technology Ministry of Education of China Hubei Wuhan China University of Texas at Arlington Department of Computer Science and Engineering ArlingtonTX United States

In large-scale information systems, storage device performance continues to improve while workloads expand in size and access characteristics. This growth puts tremendous pressure on caches and storage hierarchy in terms of concurrent throughput. However, existing cache eviction policies often struggle to provide adequate concurrent throughput due to their reliance on coarse-grained locking mechanisms and complex data structures. This paper presents a practical approach to cache eviction algorithm design, called Mobius, that optimizes the concurrent throughput of caches and reduces cache operation latency by utilizing lock-free data structures, while maintaining comparable hit ratios. Mobius includes two key designs. First, Mobius employs two lock-free FIFO queues to manage cache items, ensuring that all cache operations are executed efficiently in parallel. Second, Mobius integrates a consecutive detection mechanism that merges multiple modifications during eviction into a single operation, thereby reducing data races. Extensive evaluations using both synthetic and real-world workloads from high-concurrency clusters demonstrate that Mobius achieves a concurrent-throughput improvement ranging from 1.2× to 8.5× over state-of-the-art methods, while also maintaining lower latency and comparable cache hit ratios. The implementation of Mobius in CacheLib and RocksDB highlights its effectiveness in enhancing cache performance in practical scenarios. © 2025 ACM.

关键词： Keyways

来源：评论

学校读者我要写书评

暂无评论

Next-Generation Artificial Intelligence Solutions for Comprehensive Analysis and Improvement of Heat Dissipation in High-Intensity Boiling Environments

Next-Generation Artificial Intelligence Solutions for Compre...

引用

2024 OPJU International Technology Conference on Smart Computing for Innovation and Advancement in Industry 4.0, OTCON 2024

作者： Gadiyar, H. Manoj T. Kumar, S Vinod Raj, Vijilius Helena Yadav, Dinesh Kumar Hasoni, Mortada Kareem Sabi Lakhanpal, Sorabh Canara Engineering College Department of Computer Science and Business System India Institute of Aeronautical Engineering Hyderabad Dundigal India New Horizon College of Engineering Department of Applied Sciences Bangalore India Lloyd Institute of Engineering & Technology Knowledge Park II Uttar Pradesh Greater Noida201306 India National University of Science and Technology AlTalb Collage of Nursing Dhi Qar64001 Iraq Lovely Professional University Phagwara India

ISBN: (纸本)9798350373783

This research delves into the realm of AI-driven heat dissipation, offering a suite of algorithms to address the challenges posed by high-intensity boiling environments. The proposed predictive modeling algorithm, rooted in linear regression, provides a systematic and robust approach through 20 meticulously designed steps. This algorithm, along with neural networks, reinforcement learning, generative adversarial networks (GANs), and genetic algorithms, forms a versatile toolbox for thermal management. Linear regression, through its step-by-step predictive modeling process, ensures accurate heat dissipation predictions. Neural networks, employing backpropagation, showcase adaptability and complexity, excelling in real-time adaptation. Reinforcement learning, specifically Q-Learning, introduces dynamic decision-making strategies, vital for navigating complex thermal scenarios. GANs contribute by generating realistic synthetic data, an invaluable asset for simulating high intensity boiling environments. Genetic algorithms, resembling natural selection, prove versatile and scalable in evolving optimal solutions. The research evaluates these algorithms comprehensively, demonstrating the proposed method's superiority in prediction accuracy, energy efficiency, adaptability, and overall performance. The proposed predictive modeling algorithm stands out as a benchmark, showcasing its potential for practical implementation and optimization in diverse scenarios. This comprehensive exploration sets the stage for innovative solutions to the intricate challenges of thermal management, as continuous adaptation and fine-tuning remain pivotal in this dynamic landscape. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Text Recognition on Curved Metal Surface: An Enhanced Approach using EasyOCR and OpenCV

Text Recognition on Curved Metal Surface: An Enhanced Approa...

引用

Pervasive Computing and Social Networking (ICPCSN), International Conference on

作者： Vijayalakshmi. R A.T.N Charany S. Niresh Kumar Somu C. S Sajan. D Deepika. N Department of Computer Science and Business System Rajalakshmi Institute of Technology Chennai India Department of Information Technology Chennai Institute of Technology Chennai India Department of Computer Science and Engineering St. Joseph’s College of Engineering Chennai India Department of Computer Science and Engineering Sri Sairam Institute of Technology Chennai India Department of Computer Science and Engineering Sri Venkateswara College of Engineering Chennai India

ISBN: (数字)9798331535193

ISBN: (纸本)9798331535209

Text is a fundamental system of symbolic representation, and it plays a major role in human culture and communication. Its ability to convey the rich semantic information has become very essential in vision-based applications. With the advancement of deep learning, computer vision techniques particularly in text detection and recognition has undergone significant evolution. This paper investigates the transformative impact of deep learning on scene text detection and recognition, specifically on curved metal surfaces. By leveraging Easy Optical Character Recognition (EasyOCR) and OpenCV, the proposed approach achieves an impressive accuracy of 90% for text recognition in challenging curved surfaces. Additionally, this study explores to optimization techniques designed to enhance detection precision and examines the scalability of deep learning models in real-world applications.

关键词： Deep learning Codes Accuracy Text recognition Social networking (online) Scalability Computational modeling Text detection Metals Containers

来源：评论

学校读者我要写书评

暂无评论

Enhancing financial distress prediction through integrated Chinese Whisper clustering and federated learning

引用

Journal of Open Innovation: Technology, Market, and Complexity 2024年第3期10卷

作者： Al Ali, Amel Ibrahim S, Sheeja Rani Khedr, Ahmed M. Department of Information System University of Sharjah Department of Computer Science and Engineering American University of Sharjah Department of Computer Science University of Sharjah

Financial distress occurs when individuals or companies struggle to meet financial obligations, often due to factors like high fixed costs, illiquidity, or revenue sensitivity to economic downturns. While various machine learning and deep learning approaches have been developed for predicting financial distress by analyzing financial data to assess the likelihood of a company facing financial difficulties, challenges persist in achieving enhanced prediction accuracy and managing time complexity. To address this, a novel method known as the Chinese Whisper Clustered Stochastic Gradient Descent Federated Learning method (CWCSGDFL) is introduced. It aims to improve the efficiency of financial distress prediction by grouping data samples using the Hamann Similarity Indexed Chinese Whispers clustering process, validating them with a matching coefficient, and addressing data imbalance. Also, Kaiser-Meyer-Olkin correlative targeted projection model is employed to choose the pertinent features for prediction. By achieving balanced clustering results with these selected features, CWCSGDFL reduces time complexity in financial distress prediction. Stochastic Gradient Gaussian Kernel Federated Learning approach is employed to preserve privacy while predicting distress, aided by kriging regression for model analysis. Stochastic gradient descent momentum optimizes the global objective function, ensuring accurate classification with minimal loss. Experimental results demonstrate that CWCSGDFL achieves an average accuracy of 96%, precision of 94%, recall of 98%, F-measure of 98%, and a time complexity of 36 ms, outperforming the existing techniques. © 2024 The Authors

关键词： Financial distress prediction Hamann Similarity Indexed Chinese Whispers clustering Kriging regression Simple matching coefficient Stochastic Gradient Gaussian Kernel Federated Learning

来源：评论

学校读者我要写书评

暂无评论

Combinatorial Kinetic Dualistic Auction Model for Metaverse Service Allocation

Combinatorial Kinetic Dualistic Auction Model for Metaverse ...

引用

AITU: Digital Generation, IEEE

作者： Raj Singh Pritam Sarkar Anjan Bandyopadhyay Sujata Swain Tanvir Habib Sardar Bishwajeet Pandey School of Computer Engineering KIIT India Department of Computer Science GITAM University Bengaluru Department of Intelligent System and Cybersecurity Astana IT University Kazakhstan

ISBN: (数字)9798350364378

ISBN: (纸本)9798350364385

The focus of this research work revolves around the utilization of a specialized “Combinatorial Kinetic Dualistic Auction Model” designed specifically for “metaverse Services.” The main goal is to improve the allocation of resources in this ever-changing environment and enhance the overall experience for end-users. Through the exploration of metaverse services, we aim to further develop the comprehension and practical application of the Combinatorial Kinetic Dualistic Auction Model, ultimately contributing to a more immersive and responsive virtual experience.

关键词： Economics Greedy algorithms Technological innovation Graphical models Metaverse Market research Real-time systems

来源：评论

学校读者我要写书评

暂无评论

RADNet: Residual Attention-based Dynamic Weight Network for Microaneurysm Detection in Fundus Images: Fundus microaneurysm detection based on residual attention fusion module and dynamic weight loss 23

RADNet: Residual Attention-based Dynamic Weight Network for ...

引用

Proceedings of the 2023 4th International Symposium on Artificial Intelligence for Medicine science

作者： Yilin Han Xinpeng Zhang Fan Shi Engineering Research Center of Learning-Based Intelligent System of Ministry of Education School of Computer Science and Engineering Tianjin University of Technology China

ISBN: (纸本)9798400708138

Microaneurysms are small targets in fundus images. They are important indicators of early diabetic retinopathy. Detecting changes in the number of microaneurysms helps to judge the progression of diabetic retinopathy. However, microaneurysms occupy only a few pixels in fundus images. They have low contrast with the surrounding background, weak feature expression ability, and strong feature confusion. Detecting microaneurysms is an extremely difficult task. We propose a fundus microaneurysms detection method based on residual attention and dynamic weight. In order to achieve strong feature expression, we propose a residual attention fusion module to enhance the feature of the microaneurysms. In order to solve the problem of feature confusion between the microaneurysms and other strong interferences, we propose dynamic weight loss to constrain strong interferences at different scales. Implementation results show that our method outperforms the state-of-the-art models in all metrics on the e-ophtha-MA dataset and IDRID dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

DALR: A Dynamically Adjustable Linear Regression Model for Handling Stream Data

DALR: A Dynamically Adjustable Linear Regression Model for H...

引用

International Conference on computer, Communication, Control and Information Technology (C3IT)

作者： Rishita Mondal Sourab Mandal Paramartha Dutta Arindrajit Pal Department of Computer & System Sciences Visva-Bharati Santiniketan WB India Department of Computer Science and Engineering Academy of Technology Adisaptagram WB India

ISBN: (数字)9798350389128

ISBN: (纸本)9798350389135

In response to the escalating demand for machine learning techniques capable of handling real-time data streams, particularly in applications like stock markets, this research dives deep into the domain of stream regression. The objective is to adeptly navigate challenges posed by evolving data over time, differentiating it from conventional static regression. Existing models, such as Fast and Incremental Model Trees with Drift Detection (FIMT-DD), k-Nearest Neighbours (KNN), and Adaptive Linear Regression (ALR) and some other studies, though competent, exhibit certain limitations like learning based on certain portions of data, using a significantly large memory and not capable of handling each data in every timestamp it arrives. We propose the Dynamically Adjustable Linear Regression Model (DALR) to address these shortcomings. DALR introduces an adaptive regression model that dynamically evolves based on historical knowledge and newly acquired data, providing an innovative solution to the intricacies of dynamic data streams. This endeavor has also highlighted that the DALR algorithm consistently demonstrates strong performance across various datasets according to the RMSE metric, surpassing other state-of -the-art algorithms in most instances.

关键词： Adaptation models Heuristic algorithms Linear regression Aerodynamics Data models Robustness Real-time systems Numerical models Stock markets Streams

来源：评论

学校读者我要写书评

暂无评论

A Multi-Agent Garage Service Search and Recommendation with Hybrid MLs and LLMs

A Multi-Agent Garage Service Search and Recommendation with ...

引用

International Conference on computer, Electrical & Communication engineering (ICCECE)

作者： Mahfuzur Rahman Shuvo Ashifur Rahman Vishwanath Akuthota Tanay Paul Mahfujul Islam Md Sadi Ashraf Prosenjit Roy Md Tanzim Reza Department of Management Information System International American University United States Department of Computer Science and Engineering Chittagong University of Engineering and Technology Bangladesh TechOptima Hyderabad India Department of Computer Science and Engineering BRAC University Bangladesh Department of Computer Science Prairie View A & M University Texas United States

ISBN: (数字)9798331513894

ISBN: (纸本)9798331513900

The automobile service industry’s explosive growth highlights the need for creative approaches to boost operational effectiveness and user experience. This study introduces a Hybrid Garage Assistance system, integrating Classical Machine Learning (ML) techniques with Generative AI to optimize garage service discovery and analysis. The system employs sophisticated data processing methods, including Term Frequency-Inverse Document Frequency (TF-IDF) vectorization and regex-based service detection, to extract actionable insights from unstructured garage *** to the system are machine learning models Random Forest (RF) and XGBoost (XGB) which achieve high precision and recall in classifying garage services. A hybrid search mechanism, combining cosine similarity with ML-driven predictions, ensures the delivery of highly personalized search results. To further refine decision-making, the system incorporates Generative AI models such as Perplexity for web-based research, Gemini for location-specific analysis, Mistral for email sending and GPT-4 for detailed service recommendations and dall-e for creating user specific parts images. These advanced tools provide users with comprehensive information that enables them to make well-informed decisions about garage *** evaluation of the system is conducted using robust metrics, including precision, recall, F1-score, and system latency. Experimental results reveal a precision of $\mathbf{8 5 \%}$, recall of $\mathbf{7 0 . 8 \%}$, and an F1-score of $\mathbf{7 7 . 2 \%}$, demonstrating the efficacy of integrating classical ML with generative AI. The system’s average latency of 5.9 seconds ensures a seamless and responsive user *** hybrid framework highlights the potential of blending classical ML and Large Language Models (LLMs) to enhance search and recommendation functionalities, offering a scalable and robust blueprint for future advancements in the automotive service sector. The system’s Pro

关键词： Radio frequency Measurement Generative AI Navigation Scalability Hybrid power systems User experience Reliability Random forests Multi-agent systems

来源：评论

学校读者我要写书评

暂无评论

A Vision-Based Neural Networks Model for Turbine Trench-Filler Diagnosis 1

引用

7th International Congress and Workshop on Industrial AI and eMaintenance, IAI 2023

作者： Isaza, Cesar Guerrero-Garcia, Fernando Anaya, Karina Jenab, Kouroush Ortega-Moody, Jorge Networking and Telecommunication Academic Program Polytechnic University of Queretaro Queretaro Mexico Productive System Master Academic Program Polytechnic University of Queretaro Queretaro Mexico School of Engineering and Computer Science Morehead State University MoreheadKY40351 United States

ISBN: (数字)9783031396199

ISBN: (纸本)9783031396182

Vision-based neural networks as artificial intelligence models have been critical in many manufacturing industries, including automotive, food, and aerospace. Machine vision and deep learning have provided practical, promising, and accessible innovations to address many problems during manufacturing and diagnostics. Recent studies offer beneficial results in implementing industrial artificial intelligence systems that require determining, comparing, and evaluating optimal technological solutions. The aerospace industry has to deal with the problem of diagnosing critical components that continually need the expertise of a trained human being. On the other hand, automatic diagnostics are becoming a critical technology to deal with this problem. However, these systems require a particular configuration of computer vision algorithms which, in the case of turbine trench-filling components, have yet to be added to the literature. Considering the above, we report in this paper a new methodology that uses a pre-trained deep neural network framework available online to automatically diagnose geometric nonconformities in aeronautical components that protect the turbines from vibrations and reduce noise from the engines to the aircraft deck. The method is based on the following stages: a computer vision system with a monochrome camera to acquire images, training of deep neural networks with transfer learning, and a stage to analyze nonconformities automatically. In addition, a benchmark of several pre-trained deep neural networks is presented to address the problem. According to the experimental results, the YOLO deep neural network topology significantly contributes to automatic diagnosis with high precision and exact rate. Finally, we show that integrating deep neural networks with control graphs is a promising strategy to optimize online diagnostics in the production lines of the aeronautical industry. © The Author(s), under exclusive license to Springer Nature Switzerland

关键词： Deep neural networks

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：