检索结果-内蒙古大学图书馆

2024 International conference on Emerging Innovations and Advanced Computing, INNOCOMP 2024

作者： Santhiya, S. Jayadharshini, P. Sharmila, C. Sruthi, K. Vishnu, N.Kavin Suganth, V. Kongu Engineering College Department of Artificial Intelligence Erode India Kongu Engineering College Department of Computer Science and Engineering Erode India

ISBN: (纸本)9798350376470

Glaucoma a pervasive, incurable eye condition which diminishes eyesight and worsens the standard of life. It damages the optic nerve, which is in charge of maintaining vision in the rear of the eye and causes long-term vision loss. There are three categories in the dataset, A abnormal indicating angle-closure glaucoma and early-glaucoma indicating open-closure glaucoma and a normal eye for prediction. In this article, we formulate various deep learning models. The Convolutional Neural network (CNN) incorporates fully-connected layers at its conclusion, leveraging the extracted features from the image for the ultimate prediction. The ResNet50 model, where the numeral '50' denotes the total count of layers, encompassing both convolutional and fully connected layers. Additionally, the Multilayer Perceptron encompasses an input layer, one or more hidden layers, and an output layer. and ensembling these models to make a single model to improve the accuracy. In this ensembling method, different models can be separately trained using the same dataset (glaucoma dataset from ORIGA). Ensembling can lessen overfitting, increase robustness, and boost performance as a whole. The accuracy of the ensembling model is 89% which is more when compared to other individual models. © 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

A study of Knowledge Distillation in Fully Convolutional network for Time Series Classification

A study of Knowledge Distillation in Fully Convolutional Net...

引用

ieee International conference on Fuzzy Systems (FUZZ-ieee) / ieee World Congress on Computational Intelligence (ieee WCCI) / International Joint conference on Neural networks (IJCNN) / ieee Congress on Evolutionary Computation (ieee CEC)

作者： Ay, Emel Devanne, Maxime Weber, Jonathan Forestier, Germain Univ Haute Alsace IRIMAS Mulhouse France

ISBN: (数字)9781728186719

ISBN: (纸本)9781728186719

In recent years, deep learning revolutionized the field of machine learning. While many applications of deep learning are observed in computer vision, other domains like natural language processing (NLP) or speech recognition also benefited from advances in deep learning research. More recently, the field of time series analysis and more especially time series classification (TSC) also witnessed the emergence of deep neural networks providing competitive results. Through the years, the proposed network architectures became deeper and deeper pushing the performance higher. While these very deep models achieve impressive accuracy, their training and deployment became challenging. Indeed, a large number of GPUs is often required to train state-of-the-art networks and obtain high performances. While the requirements needed for the training step can be acceptable, deploying very deep neural networks can be difficult especially in embedded systems (e.g. robots) or devices with limited resources (e.g. web browsers, smartphones). In this context, knowledge distillation is a machine learning task consisting in transferring knowledge from a large model to a smaller one with fewer parameters. The goal is to create a lighter model mimicking the predictions of a larger one in order to obtain similar performances with a fraction of the computational cost. In this paper, we introduce and explore the concept of knowledge distillation for the specific task of TSC. We also present a first experimental study showing promising results on several datasets of the UCR time series archive. As current state-of-the-art models for TSC are deep and sometimes ensemble of models, we believe that knowledge distillation could become an important research area in the coming years.

关键词： Times Series Classification Knowledge Distillation Fully Convolutional network

来源：评论

学校读者我要写书评

暂无评论

PACT 2022 - Proceedings of the 2022 International conference on Parallel architectures and Compilation Techniques

PACT 2022 - Proceedings of the 2022 International Conference...

引用

31st International conference on Parallel architectures and Compilation Techniques, PACT 2022

ISBN: (纸本)9781450398688

The proceedings contain 50 papers. The topics discussed include: Com-CAS: effective cache apportioning under compiler guidance;transfer-tuning: reusing auto-schedules for efficient tensor program code generation;slice-and-forge: making better use of caches for graph convolutional network accelerators;GNNear: accelerating full-batch training of graph neural networks with near-memory processing;T-GCN: a sampling based streaming graph neural network system with hybrid architecture;optimizing aggregate computation of graph neural networks with On-GPU interpreter-style programming;Pavise: integrating fault tolerance support for persistent memory applications;efficient atomic durability on eADR-enabled persistent memory;custom high-performance vector code generation for data-specific sparse computations;and decoupling scheduler, topology layout, and algorithm to easily enlarge the tuning space of GPU graph processing.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Quantum Dynamic network Modeling Algorithm in Mathematical Modeling Training Combined with BOPPPS Comprehensive networking System Construction

Quantum Dynamic Network Modeling Algorithm in Mathematical M...

引用

2022 International conference on Edge Computing and Applications, ICECAA 2022

作者： Li, Mingwei Yunnan Open University Yunnan 650599 China

ISBN: (数字)9781665482325

ISBN: (纸本)9781665482325

A quantum self-organizing feature mapping network model and clustering algorithm are proposed. The inputs and weights of quantum neurons are quantum bits, and the output is real numbers. The characteristics and relationship of BOPPPS model and problem-driven teaching method are studied, BOPPPS model and teaching method The relationship between elements 5W, and taking the "Algorithm and Data Structure" course as an example, gives strategies and practices for cultivating computational thinking. The quantum self-organizing feature map network is composed of input layer and competition layer. The multi-objective programming model is established by mathematical methods such as linear weighting, objective programming and hierarchical sequence, and the solution to the dynamic allocation of resources in the logistics network is given by analyzing the model. © 2022 ieee.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

IGNNITION: Bridging the Gap between Graph Neural networks and networking Systems

引用

ieee network 2021年第6期35卷 171-177页

作者： Pujol-Perich, David Suarez-Varela, Jose Ferriol, Miquel Xiao, Shihan Wu, Bo Cabellos-Aparicio, Albert Barlet-Ros, Pere Univ Politecn Cataluna Barcelona Spain Huawei Technol Co Ltd Shenzhen Peoples R China

Recent years have seen the vast potential of graph neural networks (GNN) in many fields where data is structured as graphs (e.g., chemistry, recommender systems). In particular, GNNs are becoming increasingly popular in the field of networking, as graphs are intrinsically present at many levels (e.g., topology, routing). The main novelty of GNNs is their ability to generalize to other networks unseen during training, which is an essential feature for developing practical machine learning (ML) solutions for networking. However, implementing a functional GNN prototype is currently a cumbersome task that requires strong skills in neural network programming. This poses an important barrier to network engineers that often do not have the necessary ML expertise. In this article, we present IGNNITION, a novel open source framework that enables fast prototyping of GNNs for networking systems. IGNNITION is based on an intuitive high-level abstraction that hides the complexity behind GNNs, while still offering great flexibility to build custom GNN architectures. To showcase the versatility and performance of this framework, we implement two state-of-the-art GNN models applied to different networking use cases. Our results show that the GNN models produced by IGNNITION are equivalent in terms of accuracy and performance to their native implementations in TensorFlow.

关键词： Training data network topology Message passing Prototypes programming Routing Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Towards Pulverised architectures for Collective Adaptive Systems through Multi-tier programming 2

Towards Pulverised Architectures for Collective Adaptive Sys...

引用

2nd ieee International conference on Autonomic Computing and Self-Organizing Systems (ACSOS)

作者： Aguzzi, Gianluca Casadei, Roberto Pianini, Danilo Salvaneschi, Guido Viroli, Mirko Alma Mater Studiorum Univ Bologna Cesena Italy Univ St Gallen St Gallen Switzerland

ISBN: (纸本)9781665443937

Engineering large-scale Cyber-Physical Systems - like robot swarms, augmented crowds, and smart cities - is challenging, for many issues have to be addressed, including specifying their collective adaptive behaviour and managing the connection of the digital and physical parts. In particular, some approaches propose self-organising mechanisms to actually program global behaviour while fostering decentralised, asynchronous execution. However, most of these approaches couple behavioural specifications to specific network architectures (e.g., peer-to-peer), and therefore do not promote flexible exploitation of the underlying infrastructure. Conversely, pulverisation is a recent approach that enables self-organising behaviour to be defined independently of the available infrastructure while retaining functional correctness. However, there are currently no tools to formally specify and verify concrete architectures for pulverised applications. Therefore, we propose to combine pulverisation with multi-tier programming, a paradigm that supports the specification of the architecture of distributed systems in a single code base, and enables static checks for the correctness of actual deployments. The approach can be implemented by combining the ScaFi aggregate computing tool-chain with the ScalaLoci multi-tier programming language, paving the path to support the development of self-organising cyber-physical systems, addressing both functional (behaviour) and non-functional concerns (deployment) in a single code base and modular fashion.

关键词： Pulverisation Aggregate Computing Multi-tier programming

来源：评论

学校读者我要写书评

暂无评论

A Hybrid network of CNN and Transformer for Lightweight Image Super-Resolution

A Hybrid Network of CNN and Transformer for Lightweight Imag...

引用

ieee/CVF conference on Computer Vision and Pattern Recognition (CVPR)

作者： Fang, Jinsheng Lin, Hanjiang Chen, Xinyu Zeng, Kun Minnan Normal Univ Zhangzhou Peoples R China Minjiang Univ Fuzhou Fujian Peoples R China

ISBN: (数字)9781665487399

ISBN: (纸本)9781665487399

Recently, a number of CNN based methods have made great progress in single image super-resolution. However, these existing architectures commonly build massive number of network layers, bringing high computational complexity and heavy memory consumption, which is inappropriate to be applied on embedded terminals such as mobile platforms. In order to solve this problem, we propose a hybrid network of CNN and Transformer (HNCT) for lightweight image super-resolution. In general, HNCT consists of four parts, which are shallow feature extraction module, Hybrid Blocks of CNN and Transformer (HBCTs), dense feature fusion module and up-sampling module, respectively. By combining CNN and Transformer, HBCT extracts deep features beneficial for super-resolution reconstruction in consideration of both local and non-local priors, while being lightweight and flexible enough. Enhanced spatial attention is introduced in HBCT to further improve performance. Extensive experimental results show our HNCT is superior to the state-of-the-art methods in terms of super-resolution performance and model complexity. Moreover, we won the second best PSNR and the least activation operations in NTIRE 2022 Efficient SR Challenge. Code is available at https://***/lhjthp/HNCT.

关键词： Computer vision Codes conferences Superresolution Memory management Transformers Feature extraction

来源：评论

学校读者我要写书评

暂无评论

SemirFL: Boosting Fault Localization via Combining Semantic Information and Information Retrieval 22

SemirFL: Boosting Fault Localization via Combining Semantic ...

引用

22nd ieee International conference on Software Quality, Reliability and Security (QRS)

作者： Shi, Xiangyu Ju, Xiaolin Chen, Xiang Lu, Guilong Xu, Mengqi Nantong Univ Nantong Jiangsu Peoples R China

ISBN: (纸本)9798350319910

Automated fault localization aims to reduce software maintenance's workload during software development's evolution. Applying different features extracted from bug reports and source files can help locate faults. However, these approaches consider programming languages as natural when measuring similarity features, considering only precise term matching and ignoring deep semantic similarity features. Furthermore, existing bug localization approaches need to utilize the structural information extracted from source files, where program languages have unique structural features compared to natural languages. In this paper, we proposed SemirFL, a model combining both Convolutional Neural network(CNN) and revised Vector Space Model(rVSM), which is feeded with four metadata features (bug-fixing recency, bug-fixing frequency, collaborative filtering score, and class name similarity). SemirFL has been studied on four open-source projects. The experimental results show that SemirFL can significantly outperform the existing representative techniques in locating faults in the buggy source files.

关键词： Fault localization Deep Learning Information retrieval Convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Fast simulation of future 128-bit architectures

Fast simulation of future 128-bit architectures

引用

25th Design, Automation and Test in Europe conference and Exhibition (DATE)

作者： Portas, Fabien Petrot, Frederic Univ Grenoble Alpes CNRS Grenoble INP TIMA F-38000 Grenoble France

ISBN: (纸本)9783981926361

Whether 128-bit architectures will some day hit the market or not is an open question. There is however a trend towards that direction: virtual addresses grew from 34 to 48 bits in 1999 and then to 57 bits in 2019. The impact of a virtually infinite addressable space on software is hard to predict, but it will most likely be major. Simulation tools are therefore needed to support research and experimentation for tooling and software. In this paper, we present the implementation of the 128-bit extension of the RISC-V architecture in the QEMU functional simulator and report first performance evaluations. On our limited set of programs, simulation is slowed down by a factor of at worst 5 compared to 64-bit simulation, making the tool still usable for executing large software codes.

关键词： Performance evaluation Codes Computer architecture Market research Software

来源：评论

学校读者我要写书评

暂无评论

QoE-DASH: DASH QoE Performance Evaluation Tool for Edge-Cache and Recommendation

QoE-DASH: DASH QoE Performance Evaluation Tool for Edge-Cach...

引用

ieee International conference on Communications (ICC)

作者： Esper, Joao Paulo Bastos Loureiro Moncao, Ana Claudia Chaves Rodrigues, Karlla B. Both, Cristiano Bonato Correa, Sand Luz Cardoso, Kleber Vieira Univ Fed Goias Goiania Go Brazil Univ Vale Rio dos Sinos Sao Leopoldo Brazil

ISBN: (纸本)9781538683477

The converging ecosystem provided by Multi-access Edge Computing (MEC) has motivated novel DASH video streaming provisioning scenarios involving the joint coordination of different mechanisms for caching, communication and control. Given the complexity of designing such mechanisms, it is important to provide the research community with open-source tools that support the assessment of their feasibility, specially in real-world environment. Current network emulators still require a significant programming effort to meeting this need. To fill this gap, a new DASH emulator called QoE-DASH is presented in this work. QoE-DASH builds upon goDASH to evaluate the QoE of users consuming DASH content, taking into account network properties, user preferences, and context information. To demonstrate the capabilities of QoE-DASH, we exercise different functionalities of our tool and present a case study where two joint caching and recommendation models, proposed in the literature, are evaluated and their effects on user QoE are depicted using state-of-the-art QoE metrics.

关键词： DASH streaming service QoE MEC network emulation

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：