检索结果-内蒙古大学图书馆

15th International Conference on computing Communication and Networking Technologies, ICCCNT 2024

作者： Paul, Poulin Manishankar, S. Satheesh Kumar, R. Sahrdaya College of Engineering and Technology Dept. of Computer Science and Engineering Kerala Thrissur India Jawaharlal College of Engineering and Technology Dept. of Computer Science and Engineering Kerala Palakkad India

ISBN: (纸本)9798350370249

this paper presents an integrated framework for the comprehensive analysis of diseases affecting pomegranate fruit. the suggested system, which makes use of deep learning techniques, includes semantic segmentation for diseased region localization, multiclass illness classification, and image processing-based severity assessment. Initially, four common diseases - Alternaria, Anthracnose, Bacterial Blight, and Cer-cospora - are successfully identified through the use of a Convolutional Neural Network (CNN) for multiclass classification. then, annotated photos are used as training data for Semantic Segmentation, which uses the UNet architecture to identify the unhealthy spots within the fruit. the area and percentage of the segmented region are determined, and thresholds for high, medium, and low severity levels are then defined. the segmentation results are then used as inputs for severity estimate. Early identification and intervention tactics are made easier by the severity level prediction, which is based on the proportion of the segmented region. In order to improve pomegranate fruit management methods, the suggested framework provides a thorough method of disease analysis by combining categorization, semantic segmentation, and severity prediction. © 2024 IEEE.

关键词： Annotation Convolutional Neural Networks Deep learning Semantic segmentation Unet

来源：评论

学校读者我要写书评

暂无评论

Multi-Class SVM & Random Forest Based Intrusion Detection Using UNSW-NB15 Dataset 15

Multi-Class SVM & Random Forest Based Intrusion Detection Us...

引用

15th International Conference on computing Communication and Networking Technologies, ICCCNT 2024

作者： Jose, Anish Mathew Mukherjee, Avirup Saha, Joydeep Bhaumik, Anubhab Datta, Kamalesh Patra, Radhashyam Behura, Aradhana KIIT University BBSR School of Computer Engineering Odisha India VSSUT Department of Electronics and Telecommunication Engineering Sambalpur Odisha Burla India Universit BBSR School of Computer Engineering Odisha India

ISBN: (纸本)9798350370249

Intrusion Detection or malicious node detection (IDS) show a important role in safeguarding computer networks system from security threats. However, traditional IDS methods often struggle with high false positive rates and limited detection accuracy. this paper presents an advanced IDS that harnesses the power of a multi-class support vector machine (SVM) and integrates random forest technique for feature selection, specifically targeting the UNSW-NB15 dataset. the motivation behind this research is to enhance IDS performance by leveraging robust machine learning techniques. the research problem centers on improving the accuracy and efficiency of detecting various network attacks while reducing false positives. Our approach involves meticulously preprocessing network traffic data to extract relevant features, utilizing random forest to identify the most informative ones, and then employing these features to train a multi-class SVM classifier. this classifier effectively categorizes network traffic into different attack types. the results demonstrate a significant improvement in prediction accuracy and a decrease in rate of false positives, thereby enhancing the overall effectiveness of IDS in protecting computer networks against security threats. © 2024 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Flexible silicon photonic architecture for accelerating distributed deep learning

引用

JOURNAL OF OPTICAL COMMUNICATIONS AND NETWORKING 2024年第2期16卷 A157-A168页

作者： Wu, Zhenguo Dai, Liang Yuan Wang, Yuyang Wang, Songli Bergman, Keren Columbia Univ Dept Elect Engn 500 120th St New York NY 10027 USA

the increasing size and complexity of deep learning (DL) models have led to the wide adoption of distributed training methods in datacenters (DCs) and high-performance computing (HPC) systems. However, communication among distributed computing units (CUs) has emerged as a major bottleneck in the training process. In this study, we propose Flex-SiPAC, a flexible silicon photonic accelerated compute cluster designed to accelerate multi-tenant distributed DL training workloads. Flex-SiPAC takes a co-design approach that combines a silicon photonic hardware platform with a tailored collective algorithm, optimized to leverage the unique physical properties of the architecture. the hardware platform integrates a novel wavelength-reconfigurable transceiver design and a micro-resonator-based wavelength-reconfigurable switch, enabling the system to achieve flexible bandwidth steering in the wavelength domain. the collective algorithm is designed to support reconfigurable topologies, enabling efficient all-reduce communications that are commonly used in DL training. the feasibility of the Flex-SiPAC architecture is demonstrated through two testbed experiments. First, an optical testbed experiment demonstrates the flexible routing of wavelengths by shuffling an array of input wavelengths using a custom-designed spatial-wavelength selective switch. Second, a four-GPU testbed running two DL workloads shows a 23% improvement in job completion time compared to a similarly sized leaf-spine topology. We further evaluate Flex-SiPAC using large-scale simulations, which show that Flex-SiPAC is able to reduce the communication time by 26% to 29% compared to state-of-the-art compute clusters under representative collective operations.

关键词： Topology Optical switches Bandwidth computer architecture Training Copper Transceivers

来源：评论

学校读者我要写书评

暂无评论

Enhancing Data Security in Internet of Vehicular things (IoVT) through Blockchain-based Solutions 15

Enhancing Data Security in Internet of Vehicular Things (IoV...

引用

15th International Conference on computing Communication and Networking Technologies, ICCCNT 2024

作者： Vats, Satvik Mehta, Shiva Singh, Choudhary Ravi Graphic Era Hill University Computer Science and Engineering Dehradun248002 India Graphic Era Deemed to be University Dehradun248002 India Chitkara University Institute of Engineering and Technology Chitkara University Punjab India Invertis University Computer Science And Engineering Department Bareilly India

ISBN: (纸本)9798350370249

In the current situation, it has been confirmed that VANET, which is used to communicate between vehicles, has more capacity than MANET, which refers to conventional mobile ad-hoc networking. therefore, the ongoing development of VANETs is the fundament for enhancing the performance of traffic management systems, reducing crashes, upgrading road safety, and wisely handling emergency cases. Modern cars are composed of various new instruments that gather millions of information about features, fuel needs, and, eventually, destinations and have become the core of the global car-sharing application. through these wide storage units, sudden changes in plans by travelers and the need to go online for immediate preparations to travel are helped, which is the basic principle of ambient intelligence stipulated under the Internet of things (IoT) architecture. When there is an emergency, time is crucial. For this reason, the first important step to identifying the nearest emergency facility is to conduct a rapid examination of vehicle reports. By implementing the machine learning approach that places a high focus on the anonymity of users, this paper introduces a new method of predicting vehicles' positions based on enormous data sets. Finally, the experiment has been proven to be the most effective technique for exact vehicle position forecasting and data protection. the results clearly showed that the modeling accuracy was immensely enhanced by the prediction, which introduced a marked 12.2% improvement when compared with the previous models. Consequently, the data was calculated within 40% less time by applying the federated learning method and edge computing. Furthermore, blockchain ensures data protection and integrity, reducing the number of security vulnerabilities and attacks cases to less than 20%, making the system safer against future cyber attacks. © 2024 IEEE.

关键词： Data Privacy FedAvg Approach Internet of things Manet Security Vanet Vehicles

来源：评论

学校读者我要写书评

暂无评论

Measuring Electromagnetic performance and Configuration of Networks in high Interference Environments 15

Measuring Electromagnetic Performance and Configuration of N...

引用

15th International Conference on computing Communication and Networking Technologies, ICCCNT 2024

作者： Khan, A.Z. Panwar, Neeraj Rayaguru, N.K. Shelke, Priya M. Giri, Rakesh Kumar Jaishankar, P. Yeshwantrao Chavan College of Engineering Applied Physics Department Nagpur India University Gehu Dehradun Department of School of Computing India Vel Tech Rangarajan Dr sagunthala R & D Institute of Science and Technology Department of Eee India Vishwakarma Institute of Information Technology Department of Information Technology Pune India Sunrise University Department of Computer Science & Engineering Alwar India Sastra Deemed to be University Department of School of Civil Engineering India

ISBN: (纸本)9798350370249

Measuring the electromagnetic overall performance and configuration of networks in high-interference environments can be challenging. In those environments, the indicators of hobby can be infected with unwanted heritage indicators, i.e., noise, leading to degraded overall performance and misguided size readings. this paper investigates the most reliable techniques for measuring networks' electromagnetic performance and configuration in these problematic environments, utilizing introducing a unique multistage framework that combines sign-processing strategies with some community parameters. In particular, the framework includes four degrees: sign reception, sensitivity analysis, pass-correlation, and community configuration optimization-an analysis of the results of varying environmental conditions on sign reception and sensitivity change into conduct. A cross-correlation approach has evolved to decide the extent of correlation among one-of-a-kind alerts and to locate the specific configurations of the particular networks. Sooner or later, an optimization set of rules was carried out to locate the network's optimum configuration. the final result of this research is a reliable method for measuring the electromagnetic performance and configuration of networks in high-interference environments. the proposed technique is anticipated to help researchers and engineers better understand the electromagnetic behaviors of complicated networks in actual-world scenarios. © 2024 IEEE.

关键词： Configuration Electromagnetic Environmental performance Technique

来源：评论

学校读者我要写书评

暂无评论

Analyzing HPC Monitoring Data With a View Towards Efficient Resource Utilization 36

Analyzing HPC Monitoring Data With a View Towards Efficient ...

引用

36th IEEE International symposium on computer architecture and high-performance computing, SBAC-PAD 2024

作者： Maloney, Samuel Suarez, Estela Eicker, Norbert Guimaraes, Filipe Frings, Wolfgang Forschungszentrum Jülich Jülich Supercomputing Centre Germany University of Bonn Institute of Computer Science Germany University of Wuppertal School of Mathematics and Natural Sciences Germany

ISBN: (纸本)9798350356168

Compute nodes in modern HPC systems are growing in size and their hardware has become ever more diverse. Still, many HPC centers allocate the resources of full nodes exclusively to avoid contention, despite the associated risk of underutilization. this paper describes a thorough resource utilization study of CPU and GPU compute and memory capacity, and interconnect bandwidth on JUWELS, a mature leadership-class modular supercomputer, with the aim of identifying opportunities for improving utilization through advanced scheduling and node sharing. Separate analysis of CPU-only and GPU-accelerated nodes finds that CPU compute usage is already close to optimal for the CPU-only nodes, whereas there is plenty of scope for co-scheduling CPU-based jobs on GPU-accelerated nodes. Memory capacity and node-level interconnect bandwidth are sufficient to provision co-scheduled jobs. We analyze multiple one-month datasets to validate robustness of conclusions over time and compare with previous studies on other systems to establish generalizability of results. © 2024 IEEE.

关键词： Dynamic/Adaptive scheduling high performance computing (HPC) Monitoring Predictive analytics Resource management

来源：评论

学校读者我要写书评

暂无评论

EfficientNet-V2-B3 Transfer Learning for high-Accuracy Indian Bird Species Classification 15

EfficientNet-V2-B3 Transfer Learning for High-Accuracy India...

引用

15th International Conference on computing Communication and Networking Technologies, ICCCNT 2024

作者： Pillai, Rudresh Sharma, Neha Upadhyay, Deepak Devliyal, Swati Kaur, Gurjot Chitkara University Institute of Engineering & Technology Chitkara University Punjab India Graphic Era Hill University Computer Science & Engineering Uttarakhand Dehradun India Graphic Era Deemed to be University Computer Science &engineering Uttarakhand Dehradun India

ISBN: (纸本)9798350370249

this research paper presents a study on classifying images of Indian birds into 25 different types using the EfficientNet-V2-B3 transfer learning model. the research focuses on using deep learning techniques to accurately classify images of Indian birds, which is important for many ecological and conservation efforts. the ' Indian-Birds-Species-Image-Classification' dataset, which contains 37,500 descriptive images of 25 Indian birds, was used in the study. To train and evaluate the model, the data was split into training, validation, and testing sets. the EfficientNet-V2-B3 model was chosen for its efficiency and effectiveness in image classification, especially through transfer learning. During training, the model was trained on 24,000 images in training set and their weights were adjusted to learn the unique characteristics of Indian birds. the validation set contained 6000 images and was used to monitor the performance of the model and prevent overfitting. the pre-trained EfficientNetB3 model was adjusted throughout the training phase using stochastic gradient descent and backpropagation. Measurements like accuracy and loss were monitored throughout several epochs during the training and validation phases to get insight into the model's convergence. A confusion matrix, which provides a thorough perspective of classification accuracy and misclassifications across species, was used to examine the model's performance. the confusion matrix created from the test results provides information on performance standards for each bird species, including several performance metrics. In test phase, the model classified images of Indian birds by species with 97% accuracy. this result highlights the effectiveness of the transfer learning methodology and the EfficientNet-V2-B3 architecture's adaptability for challenging classification problems. the implications for ornithology and biodiversity preservation are profound since this automated technique streamlines the procedures f

关键词： Deep Learning EfficientNet-V2-B3 Fine-Tuning Indian Birds Species Classification Ornithology Transfer Learning

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Search Method for Accelerating Convolutional Neural architecture Search 23

A Hybrid Search Method for Accelerating Convolutional Neural...

引用

15th International Conference on Machine Learning and computing, ICMLC 2023

作者： Xun, Zhou Songbai, Liu Ka-Chun, Wong Qiuzhen, Lin Kaychen, Tan The Department of Computer Science City University of Hong Kong Hong Kong College of Computer Science and Software Engineering Shenzhen University KA-CHun Wong China College of Computer Science and Software Engineering Shenzhen University China The Department of Computing The Hong Kong Polytechnic University Hong Kong

ISBN: (纸本)9781450398411

the performance evaluation of candidate architectures is a key step in evolution-based neural architecture search (ENAS). Generally, high-fidelity evaluation is desired for finding the optimal architecture but suffers from expensive computational cost while low-fidelity evaluation can reduce search cost but may influence the search performance. In this work, a hybrid search method, HybridNAS, is proposed by synergizing these two evaluation strategies to improve the search effectiveness of ENAS. this method consists of two stages: global exploration and local exploitation. the former applies low-fidelity evaluation to explore the promising architectures efficiently. As for the latter, it exploits the most promising ones identified by the former to generate the better-performancing architectures. those two stages are iteratively repeated toward the optimal architecture. Extensive experiments on NAS-Bench-101 and NAS-Bench-201 reflect that HybridNAS can achieve comparable accuracy compared to existing popular NAS but at the expense of only 60%-80% computational cost. © 2023 ACM.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

RSQP: Problem-specific Architectural Customization for Accelerated Convex Quadratic Optimization 23

RSQP: Problem-specific Architectural Customization for Accel...

引用

50th Annual International symposium on computer architecture (ISCA)

作者： Wang, Maolin McInerney, Ian Stellato, Bartolomeo Boyd, Stephen So, Hayden Kwok-Hay AI Chip Ctr Emerging Smart Syst Hong Kong Peoples R China Univ Manchester Manchester Lancs England Princeton Univ Princeton NJ 08544 USA Stanford Univ Stanford CA 94305 USA Univ Hong Kong Hong Kong Peoples R China

ISBN: (纸本)9798400700958

Convex optimization is at the heart of many performance-critical applications across a wide range of domains. Although many high-performance hardware accelerators have been developed for specific optimization problems in the past, designing such accelerator is a challenging task and the resulting computing architecture is often so specific to the targeted application that they can hardly be reused even in a related application within the same domain. To accelerate general-purpose optimization solvers that must operate on diverse user input during run time, an ideal hardware solver should be able to adapt to the provided optimization problem dynamically while achieving high performance and power-efficiency. In this work, a hardware-accelerated general-purpose quadratic program solver, called RSQP, with reconfigurable functional units and data path that facilitate problem-specific customization is presented. RSQP uses a string-based encoding to describe the problem structure with fine granularity. Based on this encoding, functional units and datapath customized to the sparsity pattern of the problem are created by solving a dictionary-based lossless string compression problem and a mixed integer linear program respectively. RSQP has been integrated to accelerate the general-purpose quadratic programming solver OSQP and has been tested using an extensive benchmark with 120 optimization problems from 6 application domains. through architectural customization, RSQP achieves up to 7x performance improvement over its baseline generic design. Furthermore, when compared with a CPU and a GPU-accelerated implementation, RSQP achieves up to 31.2x and 6.9x end-to-end speedup on these benchmark programs, respectively. Finally, the FPGA accelerator operates at up to 6.6x lower dynamic power consumption and up to 22.7x higher power efficiency over the GPU implementation, making it an attractive solution for power-conscious datacenter applications.

关键词： convex optimization FPGA quadratic programming domain-specific architectures reconfigurable computing

来源：评论

学校读者我要写书评

暂无评论

Classification of Kidney Diseases Using Deep Neural Networks 15

Classification of Kidney Diseases Using Deep Neural Networks

引用

15th IEEE Annual computing and Communication Workshop and Conference, CCWC 2025

作者： Kashyap, Kristipati thoyajaksha Bhuvi, Moturu Sunayana, Parimi Dattu, Dudam Neeraj Chato, Lina University Of South Dakota Department of Computer Science VermillionSD United States

ISBN: (纸本)9798331507695

Kidney-related diseases such as tumors, cysts, and stones are common, so prompt and precise diagnosis is essential for better patient outcomes. In this work, we propose developing an automatic Deep Learning (DL) model based on various DL structures and learning techniques to create an efficient prediction model for diagnosis of kidney diseases and disorders. A publicly available dataset of computer Tomography (CT) scan images with three kidney disease categories (tumor, cyst, and stone), as well as the normal category, are used to test our proposed DL multi-classification structures: Convolutional Neural Network (CNN), CNN with transfer learning, Vision Transformer (ViT) DL based model, and ViT transfer learning-based model. Transfer learning from state-of-the-art pre-trained DL models, including VGG16, ResNet50, and ViT, are used to evaluate its effectiveness in improving performance and reducing the dependency on large training datasets. Preprocessing and data augmentation are applied to enhance image quality and ensure reliable model performance. the Vision Transformer with transfer learning achieved a high test accuracy of 99.6% on test data, with accuracy improvements of 1.13% and 1.46% compared to CNN transfer learning based and CNN classical learning based, respectively. © 2025 IEEE.

关键词： Convolutional 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：