检索结果-内蒙古大学图书馆

A multi-feature-based intelligent redundancy elimination scheme for cloud-assisted health systems

CAAI Transactions on Intelligence Technology 2024年第2期9卷 491-510页

作者： Ling Xiao Beiji Zou Xiaoyan Kui Chengzhang Zhu Wensheng Zhang Xuebing Yang Bob Zhang School of Computer Science and Engineering Central South UniversityChangshaChina Hunan Engineering Research Center of Machine Vision and Intelligent Medicine Central South UniversityChangshaChina The College of Literature and Journalism Central South UniversityChangshaChina Institute of Automation Chinese Academy of SciencesBeijingChina Department of Computer and Information Science University of MacaoMacaoChina

Redundancy elimination techniques are extensively investigated to reduce storage overheads for cloud-assisted health *** eliminates the redundancy of duplicate blocks by storing one physical instance referenced by multiple *** compression is usually regarded as a complementary technique to deduplication to further remove the redundancy of similar blocks,but our observations indicate that this is disobedient when data have sparse duplicate *** addition,there are many overlapped deltas in the resemblance detection process of post-deduplication delta compression,which hinders the efficiency of delta compression and the index phase of resemblance detection inquires abundant non-similar blocks,resulting in inefficient system ***,a multi-feature-based redundancy elimination scheme,called MFRE,is proposed to solve these *** similarity feature and temporal locality feature are excavated to assist redundancy elimination where the similarity feature well expresses the duplicate ***,similarity-based dynamic post-deduplication delta compression and temporal locality-based dynamic delta compression discover more similar base blocks to minimise overlapped deltas and improve compression ***,the clustering method based on block-relationship and the feature index strategy based on bloom filters reduce IO overheads and improve system *** demonstrate that the proposed method,compared to the state-of-the-art method,improves the compression ratio and system throughput by 9.68%and 50%,respectively.

关键词： big data cloud computing compression data compression medical applications performance evaluation

来源：评论

学校读者我要写书评

暂无评论

Hydra: bidirectional state space models through generalized matrix mixers 24

Hydra: bidirectional state space models through generalized ...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Sukjun Hwang Aakash Lahoti Ratish Puduppully Tri Dao Albert Gu Machine Learning Department Carnegie Mellon University IT University of Copenhagen Department of Computer Science Princeton University Machine Learning Department Carnegie Mellon University and Cartesia AI

ISBN: (纸本)9798331314385

A wide array of sequence models are built on a framework modeled after Transformers, comprising alternating sequence mixer and channel mixer layers. This paper studies a unifying matrix mixer view of sequence mixers that can be conceptualized as a linear map on the input sequence. This framework encompasses a broad range of well-known sequence models, including the self-attention of Transformers as well as recent strong alternatives such as structured state space models (SSMs), and allows understanding downstream characteristics such as efficiency and expressivity through properties of their structured matrix class. We identify a key axis of matrix parameterizations termed sequence alignment, which increases the flexibility and performance of matrix mixers, providing insights into the strong performance of Transformers and recent SSMs such as Mamba. Furthermore, the matrix mixer framework offers a systematic approach to developing sequence mixers with desired properties, allowing us to develop several new sub-quadratic sequence models. In particular, we propose a natural bidirectional extension of the Mamba model (Hydra), parameterized as a quasiseparable matrix mixer, which demonstrates superior performance over other sequence models including Transformers on non-causal tasks. As a drop-in replacement for attention layers, Hydra outperforms BERT by 0.8 points on the GLUE benchmark and ViT by 2% Top-1 accuracy on ImageNet.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Prompting for LLM-Based Generative Internet of Things

引用

IEEE Internet of Things Journal 2025年第1期12卷 778-791页

作者： Xiao, Bin Kantarci, Burak Kang, Jiawen Niyato, Dusit Guizani, Mohsen University of Ottawa School of Electrical Engineering and Computer Science OttawaONK1N 6N5 Canada Guangdong University of Technology School of Automation Guangzhou510006 China Nanyang Technological University School of Computer Science and Engineering Jurong West Singapore Singapore Mohamed bin Zayed University of Artificial Intelligence Machine Learning Department Abu Dhabi United Arab Emirates

Large language models (LLMs) have demonstrated remarkable capacities on various tasks, and integrating the capacities of LLMs into the Internet of Things (IoT) applications has drawn much research attention recently. Due to security concerns, many institutions avoid accessing state-of-the-art commercial LLM services, requiring the deployment and utilization of open-source LLMs in a local network setting. However, open-source LLMs usually have more limitations regarding their performance, such as their arithmetic calculation and reasoning capacities, and practical systems of applying LLMs to IoT have yet to be well-explored. Therefore, we propose an LLM-based Generative IoT (GIoT) system deployed in the local network setting in this study. To alleviate the limitations of LLMs and provide service with competitive performance, we apply prompt engineering methods to enhance the capacities of the open-source LLMs, design a Prompt Management Module and a Postprocessing Module to manage the tailored prompts for different tasks and process the results generated by the LLMs. To demonstrate the effectiveness of the proposed system, we discuss a challenging table question answering (Table-QA) task as a case study of the proposed system, as tabular data is usually more challenging than plaintext because of their complex structures, heterogeneous data types and sometimes huge sizes. We conduct comprehensive experiments on the two popular Table-QA data sets, and the results show that our proposal can achieve competitive performance compared with state-of-the-art LLMs, demonstrating that the proposed LLM-based GIoT system can provide competitive performance with tailored prompting methods and is easily extensible to new tasks without training. © 2014 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Understanding and minimising outlier features in transformer training 24

Understanding and minimising outlier features in transformer...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Bobby He Lorenzo Noci Daniele Paliotta Imanol Schlag Thomas Hofmann Department of Computer Science ETH Zürich Machine Learning Group University of Geneva

ISBN: (纸本)9798331314385

Outlier Features (OFs) are neurons whose activation magnitudes significantly exceed the average over a neural network's (NN) width. They are well known to emerge during standard transformer training and have the undesirable effect of hindering quantisation in afflicted models. Despite their practical importance, little is known behind why OFs emerge during training, nor how one can minimise them. Our work focuses on the above questions, first identifying several quantitative metrics, such as the kurtosis over neuron activation norms, to measure OFs. With these metrics, we study how architectural and optimisation choices influence OFs, and provide practical insights to minimise OFs during training. As highlights, we introduce a novel unnormalised transformer block, the Outlier Protected block, and present a previously unknown benefit of non-diagonal preconditioning optimisers, finding both approaches to significantly reduce OFs and improve quantisation without compromising convergence speed, at scales of up to 7B parameters. Notably, our combination of OP block and non-diagonal preconditioner (SOAP) achieves 14.87 weight-and-activation int8 perplexity (from 14.71 in standard precision), compared to 63.4 int8 perplexity (from 16.00) with a default OF-prone combination of Pre-Norm model and Adam, when quantising OPT-125m models post-training.

关键词：

来源：评论

学校读者我要写书评

暂无评论

arXiv

引用

arXiv 2024年

作者： Meister, Clara Giulianelli, Mario Pimentel, Tiago ETH Zürich Department of Computer Science Institute for Machine Learning Switzerland

Surprisal theory posits that the cognitive effort required to comprehend a word is determined by its contextual predictability, quantified as surprisal. Traditionally, surprisal theory treats words as distinct entities, overlooking any potential similarity between them. Giulianelli et al. (2023) address this limitation by introducing information value, a measure of predictability designed to account for similarities between communicative units. Our work leverages Ricotta and Szeidl’s (2006) diversity index to extend surprisal into a metric that we term similarity-adjusted surprisal, exposing a mathematical relationship between surprisal and information value. Similarity-adjusted surprisal aligns with information value when considering graded similarities and reduces to standard surprisal when words are treated as distinct. Experimental results with reading time data indicate that similarity-adjusted surprisal adds predictive power beyond standard surprisal for certain datasets, suggesting it serves as a complementary measure of comprehension effort. Copyright © 2024, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Blockchain for Secure and Efficient Crowdfunding: An Optimized Particle Swarm Approach 9

Blockchain for Secure and Efficient Crowdfunding: An Optimiz...

引用

9th International Conference on Communication and Electronics Systems, ICCES 2024

作者： Lakshmanan, M. Mala, G.S. Anandha Poorni, R. Ilamurugan, G. Sriramkumar, R. Gnanavel, R. Dayananda Sagar University Department of Artificial Intelligence and Machine Learning Bangalore India Easwari Engineering College Anna Univeristy Department of Computer Science and Engineering Chennai India Sri Venkateswara College of Engineering Sriperumbudur India

ISBN: (纸本)9798350377972

Crowdfunding, a method of raising capital through collective contributions, often faces challenges related to transaction efficiency, security, and transparency. Blockchain addresses these issues by providing a decentralized, transparent, and secure environment for transactions, while smart contracts automate the fund release process based on predefined conditions. PSO, a nature-inspired optimization algorithm, is applied to optimize various crowdfunding parameters such as funding strategies, donor behavior, and campaign success. By combining PSO with blockchain, the proposed model improves both the speed and security of crowdfunding transactions. This research presents an innovative approach to enhancing crowdfunding platforms by integrating Particle Swarm Optimization (PSO) with blockchain technology. Proposed Model achieves a 30% reduction in transaction time and a 40% improvement in security. This is achieved through the use of a Merkle tree for secure data storage and smart contracts to ensure efficient and automated fund management. The integration of PSO further optimizes key parameters, such as fund distribution and reward allocation, leading to a more efficient and transparent platform. The combination of these technologies results in a crowdfunding experience that is not only faster but also more secure and reliable, enabling global, borderless funding while minimizing fraud and enhancing trust among participants. This research highlights the potential of OPSA-enabled blockchain solutions to transform the crowdfunding landscape, offering a more efficient, secure, and transparent funding platform. © 2024 IEEE.

关键词： Decentralized finance

来源：评论

学校读者我要写书评

暂无评论

Hybrid Approach of CNN & RNN for Breast Cancer Detection and Segmentation 3

Hybrid Approach of CNN & RNN for Breast Cancer Detection and...

引用

3rd International Conference on Smart Technologies and Systems for Next Generation Computing, ICSTSN 2024

作者： Manju Bala, P. Priyadharshini, S. Usharani, S. Balachandar, A. IFET College of Engineering An Autonomous Institution Department of Artificial Intelligence and Data Science India IFET College of Engineering An Autonomous Institution Department of Computer Science and Engineering India IFET College of Engineering An Autonomous Institution Department of Artificial Intelligence and Machine Learning India

ISBN: (纸本)9798350391565

Breast cancer is a serious worldwide health concern, and advanced diagnostic tools are needed for an accurate and timely identification of the illness. In order to classify breast cancer through ultrasound pictures, this method suggests a hybrid strategy that combines Convolutional Neural Networks (CNNs) with Recurrent Neural Networks (RNNs). CNN integration makes it possible to extract spatial data from ultrasound pictures and identify complex patterns that may be signs of cancer. Concurrently, RNNs are used to record temporal relationships between successive ultrasound frames, which enables the model to identify changing features of possible anomalies. By combining the advantages of both CNNs & RNNs, the suggested hybrid architecture improves the model's capacity to identify minute, dynamic patterns that are essential for precise diagnosis. In comparison to standalone CNN or RNN models, the hybrid model performs better in classification, as evidenced by experimental assessments carried out on a variety of ultrasound picture datasets. The findings highlight the hybrid approach's potential to improve breast cancer diagnostic skills and present a viable path for raising the precision and dependability of ultrasound-based tumour classification techniques. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

IoT And IIoT Integrated Spruce Grid System

IoT And IIoT Integrated Spruce Grid System

引用

2023 International Conference on Power Energy, Environment and Intelligent Control, PEEIC 2023

作者： Gupta, Anand Kumar Raghavendra, R. Sikri, Alisha Tula's Institute Department of Computer Science and Engineering Dehradun India Department of Computer Bangalore India Noida Institute of Engineering and Technology Department of Artificial Intelligence & Machine Learning Uttar Pradesh Greater Noida India

ISBN: (纸本)9798350357769

The Network of Things is a rapidly developing field of technology that provides creative solutions in many different areas, most notably essential infrastructures. Its revolutionary effect is especially seen in the field of standard electricity systems, where it makes the transition to a more intelligent and efficient energy grid easier. This analysis explores the architecture and features of Internet of Things (IoT)-enabled intelligent energy grid systems, with an emphasis on several IoT technologies (sensing, communication, and computation) and their respective standards within the smart energy grid context. This page also offers a thorough summary of previous studies on the use of IoT in systems with smart grids. A review of current studies and literature indicates a significant worry about the security flaws that IoT technologies and energy systems are known to have. As a result, we examine popular threat as attack models that are unique to IoT-enabled power plants and compile mitigation strategies designed to deal with these security issues. The article's conclusion highlights the use of cutting-edge technology, such as blockchain, artificial intelligence, and machine learning, to improve the security and resilience of IoT-enabled energy networks. These cutting-edge solutions help to improve the efficacy, stability, and dependability of smart energy systems by tackling contemporary issues including security flaws. This comprehensive investigation offers insightful information on the architecture of IoT-enabled clever energy systems, the security issues they raise, and how cutting-edge technology might improve their overall efficacy systems. © 2023 IEEE.

关键词： Advanced Metering Infrastructure (AMI) cybersecurity energy system optimization fog computing and SCADA technologies Internet of Things (IoT) technologies IoT-enabled smart energy grid renewable energy integration self-healing mechanisms smart grid architecture smart meters

来源：评论

学校读者我要写书评

暂无评论

Classification of Batik Industry Wastewater Quality Using machine learning and IoT 6

Classification of Batik Industry Wastewater Quality Using Ma...

引用

6th International Seminar on Research of Information Technology and Intelligent Systems, ISRITI 2023

作者： Aldiki Febriantono, M. Zunaidah, Asih School of Computer Science Bina Nusantara University Computer Science Department Jakarta Indonesia BINUS University Digital Language Learning Center Communication Science Faculty of Humanities Jakarta Indonesia

ISBN: (纸本)9798350358346

Batik industrial wastewater contains various chemical substances that can pollute the environment and threaten human health. Therefore, it is important to monitor and analyze wastewater quality efficiently and accurately. In this study, we collect data on the quality of batik industrial wastewater using IoT technology. The data includes parameters such as pH, BOD (Biochemical Oxygen Demand), COD (Chemical Oxygen Demand), temperature, and other heavy metal concentrations. The data is then processed and prepared for the classification process. machine learning methods, especially classification algorithms, are implemented to classify wastewater quality into several categories based on predetermined quality standards. The three machine learning algorithms used in this research are SVM (Support Vector machine), KNN (K-Nearest Neighbors), and DT (Decision Tree). The results of this study indicate that the use of the SVM, KNN, and DT algorithms in classifying the quality of batik industry wastewater produces a high degree of accuracy. This system can help identify potential problems in waste management more effectively. As such, this research makes an important contribution in the development of technological solutions to address environmental problems associated with the batik industry and has the potential to become a foundation for the development of better wastewater quality monitoring and analysis systems. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

DSG: An End-to-End Document Structure Generator

DSG: An End-to-End Document Structure Generator

引用

IEEE International Conference on Data Mining (ICDM)

作者： Johannes Rausch Gentiana Rashiti Maxim Gusev Ce Zhang Stefan Feuerriegel Department of Computer Science ETH Zurich Munich Center for Machine Learning LMU Munich

Information in industry, research, and the public sector is widely stored as rendered documents (e.g., PDF files, scans). Hence, to enable downstream tasks, systems are needed that map rendered documents onto a structured hierarchical format. However, existing systems for this task are limited by heuristics and are not end-to-end trainable. In this work, we introduce the Document Structure Generator (DSG), a novel system for document parsing that is fully end-to-end trainable. DSG combines a deep neural network for parsing (i) entities in documents (e.g., figures, text blocks, headers, etc.) and (ii) relations that capture the sequence and nested structure between entities. Unlike existing systems that rely on heuristics, our DSG is trained end-to-end, making it effective and flexible for real-world applications. We further contribute a new, large-scale dataset called E-Periodica comprising real-world magazines with complex document structures for evaluation. Our results demonstrate that our DSG outperforms commercial OCR tools and, on top of that, achieves state-of-the-art performance. To the best of our knowledge, our DSG system is the first end-to-end trainable system for hierarchical document parsing.

关键词：

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：