检索结果-内蒙古大学图书馆

22nd IEEE International Symposium on Parallel and Distributed Processing with Applications, ISPA 2024

作者： Zhang, Zhi Zhang, Shiwen Chen, Shuang Liang, Wei Hunan University of Science and Technology College of Computer Science and Technology Xiangtan China Hunan University of Science and Technology Hunan Key Laboratory for Service Computing and Novel Software Technology School of Computer Science and Engineering Xiangtan411201 China

ISBN: (纸本)9798331509712

In the era of big data and growing privacy concerns, Federated Learning (FL) has emerged as a promising solution for collaborative model training while preserving user data privacy. However, FL faces challenges such as inconsistent data quality and frequent client dropouts, which hinder its performance. To address these challenges, we propose FedHPM, a novel framework that leverages the predictive capabilities of Long Short-Term Memory (LSTM) networks. FedHPM incorporates an intelligent client selection mechanism that utilizes LSTM-based analysis of historical data to forecast clients' potential data quality and dropout likelihood. This mechanism enables efficient selection of the top best-performing clients for participation in each round of global model training. By mitigating the impact of dropouts and optimizing the composition of participating clients, FedHPM significantly improves both the convergence speed and final performance of the global model. Through extensive empirical evaluations, we demonstrate the superior performance of FedHPM in enhancing FL stability, accelerating the convergence rate, and maintaining model accuracy. The results highlight the effectiveness of our framework in addressing the challenges associated with data quality and dropouts, thereby advancing the field of federated learning. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Knowledge-Guided Cross-Topic Visual Question Generation 30

Knowledge-Guided Cross-Topic Visual Question Generation

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Liu, Hongfei Wang, Guohua Xie, Jiayuan Chen, Jiali Fang, Wenhao Cai, Yi School of Software Engineering South China University of Technology China Key Laboratory of Big Data and Intelligent Robot South China University of Technology Ministry of Education China Department of Computing The Hong Kong Polytechnic University Hong Kong

ISBN: (纸本)9782493814104

Visual question generation (VQG) task aims to generate high-quality questions based on the input image. Current methods primarily focus on generating questions containing specified content utilizing answers or question types as constraints. However, these constraints make it challenging to control the topic of generated questions (e.g., conversation or test subject topics) for various applications. Thus, it is necessary to utilize topics as constraints to guide question generation. Considering that there are many topics and it is almost impossible for human annotations to cover them, we propose the cross-topic learning VQG (CTL-VQG) task, which aims to generate questions related to unseen topics in cross-topic scenarios. In this paper, we propose a knowledge-guided cross-topic visual question generation (KC-VQG) model to extract unseen topic-related information for question generation. Specifically, an image-topic feature extractor is introduced in our model to extract topic-related intuitive visual features;an image-topic knowledge extractor is used to extract and select the most appropriate topic-related implicit knowledge from large language models for generating questions. Extensive experiments show that our model outperforms baselines and can effectively generate unseen topic-related questions in cross-topic scenarios. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

NCL: Textual Backdoor Defense Using Noise-Augmented Contrastive Learning

NCL: Textual Backdoor Defense Using Noise-Augmented Contrast...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Shengfang Zhai Qingni Shen Xiaoyi Chen Weilong Wang Cong Li Yuejian Fang Zhonghai Wu School of Software and Microelectronics & National Engineering Research Center for Software Engineering & PKU-OCTA Laboratory for Blockchain and Privacy Computing Peking University

At present, backdoor attacks attract attention as they do great harm to deep learning models. By poisoning the training data, the adversary makes the model trained based on this dataset being injected with a backdoor. In the field of text, however, existing works do not provide sufficient defense against backdoor attacks. In this paper, we propose a Noise-augmented Contrastive Learning (NCL) framework to defend against textual backdoor attacks when training models with untrustworthy data. With the aim of mitigating the mapping between triggers and the target label, we add appropriate noise perturbing possible backdoor triggers, augment the training dataset, and then pull homology samples in the feature space utilizing contrastive learning objective. Experiments demonstrate the effectiveness of our method in defending three types of textual backdoor attacks, outperforming the prior works.

关键词： Training Deep learning Training data Signal processing Data models Acoustics Speech processing

来源：评论

学校读者我要写书评

暂无评论

A Fast Scalable Hardware Priority Queue and Optimizations for Multi-Pushes

A Fast Scalable Hardware Priority Queue and Optimizations fo...

引用

IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW)

作者： Samuel Collinson Allan Bai Oliver Sinnen Department of Electrical Computer and Software Engineering Parallel and Reconfigurable Computing Lab University of Auckland New Zealand

ISBN: (数字)9798350364606

ISBN: (纸本)9798350364613

Priority queues (PQ) are an essential data structure for many important algorithms. Hardware implementations of priority queues can accelerate such algorithms significantly. Usually a choice has to be made between very fast (fixed-size) hardware PQs or scalable PQs. Fast hardware queues can provide single cycle push and pop operations, but are limited in size. Scalable hardware queues use embedded memory to achieve scalability, but thereby compromise on speed. This paper proposes a fast and scalable hardware priority queue that can be used for general purpose. It is based on a modular and flexible hybrid design, using a shift register queue combined with a heap-based queue. In the best case, i.e. when only the first queue is needed, push and pop operations only take a single cycle. In an experimental evaluation on a Xilinx FPGA we demonstrate that performance of the hybrid queue maintains a good balance between performance and scalability even for applications where the size of the working set of data can have large variance. We further propose an optimisation of the heap-based queue to support workloads that repeat several consecutive push operations followed by a single pop operation, which is the workload, for instance, in state space search or ray-tracing.

关键词： Distributed processing Scalability Sociology Ray tracing Shift registers Data structures Hardware

来源：评论

学校读者我要写书评

暂无评论

Revolutionizing Image Recognition: Next-Generation CNN Architectures for Handwritten Digits and Objects 8

Revolutionizing Image Recognition: Next-Generation CNN Archi...

引用

8th IEEE Symposium on Wireless Technology and Applications, ISWTA 2024

作者： Absur, Md Nurul Nasif, Kazi Fahim Ahmad Saha, Sourya Nova, Sifat Nawrin Department of Computer Science City University of New York New York United States College of Computing and Software Engineering Kennesaw State University GA United States Department of Computer Science Chalmers University of Technology Gothenburg Sweden

ISBN: (纸本)9798350351354

This study addresses the pressing need for computer systems to interpret digital media images with a level of sophistication comparable to human visual perception. By leveraging Convolutional Neural Networks (CNNs), we introduce two innovative architectures tailored to distinct datasets: the MNIST handwritten digit dataset and the Fashion MNIST dataset. Unlike traditional machine learning methods such as Support Vector Machines (SVM) and Random Forests, our customized CNN models remarkably enhance image attribute comprehension and recognition accuracy. Specifically, the model developed for the MNIST dataset achieved an unprecedented accuracy of 98.71% without any bias, while the Fashion MNIST model reached 91.39%, marking significant advancements over conventional algorithms without any bias. This research showcases the superior efficiency of CNNs in processing and understanding digital images. It underscores the potential of deep learning technologies in bridging the gap between computational systems and human-like visual recognition. Through meticulous experimentation and analysis, we illustrate how deep CNNs require less preparatory work than other image-processing algorithms, setting a new benchmark in computer vision. © 2024 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Hadiyyissa Automatic Speech Recognition Using Deep Learning Approach 5

Hadiyyissa Automatic Speech Recognition Using Deep Learning ...

引用

5th International Conference on Information and Communication Technology for Development for Africa, ICT4DA 2023

作者： Shamore, Selamu Gure, Amin Tuni Yimer, Mohammed Abebe Wachemo University Department of Computer Science Hossana Ethiopia Arsi University Department of Information System Arsi Ethiopia Arba Minch University Faculty of Computing and Software Engineering Arba Minch Ethiopia

ISBN: (纸本)9798350328486

Speech Recognition focuses on developing advanced technologies that can recognize and translate spoken language into text. Automatic Speech Recognition (ASR) is an Artificial Intelligence (AI) technology that captures human voice from a mi-crophone and analyzes it based on spoken words to comprehend natural language. In Ethiopia, ASR models have been developed for specific local languages such as Amharic, Tigrigna, Wolaytta, and Afaan Oromo, but there has been a lack of research on ASR models for the Hadiyyisa language, which is also spoken in Ethiopia. The research study designed and developed an ASR model for the Hadiyyisa language. The researchers collected data from native speakers of the Hadiyyisa language and used Python programming to design and develop the deep learning model for the Hadiyissa ASR model. Different methods and techniques were used in the study to achieve the research objectives and complete the entire research process. The proposed model utilized deep learning algorithms including RNN, LSTM, BiLSTM, and CNN, along with suitable feature extraction techniques such as Mel Frequency Cepstral Coefficients (MFCC). The accuracy of the proposed ASR model was evaluated through experimentation, which demonstrated that the BiLSTM model outperformed the other deep learning models with an overall accuracy of 90.95 % Based on the experimental evidence, the proposed model was found to be better suited for ASR for the Hadiyyisa language. © 2023 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

The Ontology-Based Semantic Reasoning Method for Non-standard Part Datum Construction

The Ontology-Based Semantic Reasoning Method for Non-standar...

引用

UNIfied Conference of International Workshop on Defence Applications of Multi-Agent Systems, DAMAS 2023, International Conference on Maintenance engineering, IncoME-V 2023, International conference on the Efficiency and Performance engineering Network, TEPEN 2023

作者： Kong, Chao Xu, Yuanping Zhu, Yue Xu, Zhijie Jin, Jin Guo, Benjun Li, Tukun School of Software Engineering Chengdu University of Information Technology Chengdu China School of Computing and Engineering University of Huddersfield Queensgate Huddersfield United Kingdom China Productivity Center for Machinery Co. Ltd. Haidian District No. 2 Shouti South Road Beijing China

ISBN: (纸本)9783031494123

In the field of mechanical engineering, standard parts, such as shaft and gear, have standardized tolerance specification design process due to its formalized shapes and dimensions. However, non-standard parts are much more different because their geometrical shape and size are not fixed. Non-standard parts have widely application in automotive and aerospace, and the tolerance specification is mainly dependent on engineer experience and default standards. To this end, this study proposed an ontology-based method of datum reference frame (DRF) for non-standard parts. In this study, a hierarchical tolerance information representation model was constructed. The knowledge reasoning algorithm was established via experience and ISO GPS standards, considering the knowledge isomerism of human engineers, a CAE model was proposed to evaluate the reliability of reasoning results. A case study was conducted to demonstrate the effectiveness of proposed method, the algorithm presented in this paper can be used in the computer-aided tolerance development. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Ontology

来源：评论

学校读者我要写书评

暂无评论

A Code Completion Approach Combining Pointer Network and Transformer-XL Network 19th

A Code Completion Approach Combining Pointer Network and Tr...

引用

19th EAI International Conference on Collaborative computing: Networking, Applications and Worksharing, CollaborateCom 2023

作者： Zhang, Xiangping Liu, Jianxun Long, Teng Hu, Haize School of Computer Science and Engineering Hunan University of Science and Technology Hunan Xiangtan411100 China Hunan Key Lab. for Services Computing and Novel Software Technology Hunan University of Science and Technology Hunan Xiangtan411100 China

ISBN: (纸本)9783031545207

Code completion is an integral component of modern integrated development environments, as it not only facilitates the software development process but also enhances the quality of software products. By leveraging large-scale codes to learn the probability distribution among code token units, deep learning methods have demonstrated significant improvements in the accuracy of token unit recommendations. However, the effectiveness of code completion with deep learning techniques is hindered by information loss. To alleviate the above problem, we proposed a code language model which combines the pointer network and Transformer-XL network to overcome the limitations of existing approaches in code completion. The proposed model takes as input the original code fragment and its corresponding abstract syntax tree and leverages the Transformer-XL model as the basis model for capturing long-term dependencies. Furthermore, we integrate a pointer network as a local component to predict the out-of-vocabulary words. The proposed method is evaluated on real PY150 and JS150 datasets. The comparative experimental results demonstrate the effectiveness of our model in improving the accuracy of the code completion task at the token unit level. © ICST Institute for Computer Sciences, Social Informatics and Telecommunications engineering 2024.

关键词： Probability distributions

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Algorithms for Gold Price Prediction

Machine Learning Algorithms for Gold Price Prediction

引用

作者： Nguyen, Duyen Mai Thi Debnath, Narayan C. Quach, Luyl-Da Nguyen, Vinh Dinh School of Computing and Information Technology Eastern International University Thủ Dầu Một Viet Nam Department of Software Engineering FPT University Can Tho City Viet Nam

Due to its potential for reliably anticipating gold prices, machine learning for gold price prediction has become a hot study topic in recent years, which is essential for investors and policymakers. This research did a comprehensive survey and experiments on various machine learning algorithms, such as linear regression, support vector machines, decision trees, gaussian process regression, random forest regression, polynomial features, lasso regression, k-nearest neighbor, ridge regression and neural networks, to gold price prediction. These models have been able to provide accurate predictions of gold prices, often outperforming traditional statistical methods. However, the quality of the input data and the model selection procedure affect how accurate these models are. Comprehensive experimental results showed that the support vector machine algorithm achieved the most accurate predictions, with the lowest error rates and the best R2 score. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Promise and Peril of Collaborative Code Generation Models: Balancing Effectiveness and Memorization

arXiv

引用

arXiv 2024年

作者： Chen, Zhi Jiang, Lingxiao Centre for Research on Intelligent Software Engineering School of Computing and Information Systems Singapore Management University Singapore

In the rapidly evolving field of machine learning, training models with datasets from various locations and organizations presents significant challenges due to privacy and legal concerns. The exploration of effective collaborative training settings, which are capable of leveraging valuable knowledge from distributed and isolated datasets, is increasingly *** study investigates key factors that impact the effectiveness of collaborative training methods in code next-token prediction, as well as the correctness and utility of the generated code, showing the promise of such methods. Additionally, we evaluate the memorization of different participant training data across various collaborative training settings, including centralized, federated, and incremental training, showing their potential risks in leaking data. Our findings indicate that the size and diversity of code datasets are pivotal factors influencing the success of collaborative trained code models. We demonstrate that federated learning achieves competitive performance compared to centralized training while offering better data protection, as evidenced by lower memorization ratios in the generated code. However, federated learning can still produce verbatim code snippets from hidden training data, potentially violating data privacy or copyright. Our study further explores the patterns of effectiveness and memorization in incremental learning, emphasizing the importance of the sequence in which individual participant datasets are introduced. Also, we identify the memorization phenomenon of cross-organizational clones as a prevalent challenge in both centralized and federated learning scenarios. Our findings highlight the persistent risk of data leakage during inference, even when training data remains unseen. We conclude with strategic recommendations for practitioners and researchers to optimize the use of multisource datasets, thereby propelling the cross-organizational collaboration forward. Cop

关键词： Collaborative 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：