检索结果-内蒙古大学图书馆

Toward actionable testing of deep learning models

Science China(Information Sciences) 2023年第7期66卷 300-302页

作者： Yingfei XIONG Yongqiang TIAN Yepang LIU Shing-Chi CHEUNG Key Laboratory of High Confidence Software Technologies(Peking University) Ministry of Education School of Computer Science Peking University Department of Computer Science and Engineering The Hong Kong University of Science and Technology School of Computer Science University of Waterloo Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology Department of Computer Science and Engineering Southern University of Science and Technology

Deep learning has become an important computational paradigm in our daily lives with a wide range of applications,from authentication using facial recognition to autonomous driving in smart vehicles. The quality of the deep learning models, i.e., neural architectures with parameters trained over a dataset, is crucial to our daily living and economy.

关键词：

来源：评论

学校读者我要写书评

暂无评论

HTTP/2 Web Performance Characteristics with Link-aggregation Mode Under DDoS Attack 12

HTTP/2 Web Performance Characteristics with Link-aggregation...

引用

12th International Electrical engineering Congress, iEECON 2024

作者： Rattanaopas, Kritwara Wattanakul, Phantiga Prince Of Songkla University Faculty Of Engineering Department Of Computer Engineering Songkhla Thailand Business Computer Program Faculty Of Management Sciences Nakhon Pathom Rajabhat University Nakhon Pathom Thailand

ISBN: (纸本)9798350383591

The HTTP/2 Web server is widely used in Internet service. Furthermore, attacks known as distributed denial-of service, or DDoS, were popping up online. HTTP/2 not only allows multiple requests per connection but also protects against DDoS attacks. This paper focused on network link aggregation for bonding multiple Gigabit Ethernets of an HTTP/2 physical machine. Also, we used four link-aggregation modes with no special network switch support, including Round-Robin (Mode 0), Load Balance (balance-xor) (Mode 2), Balance-TLB (Mode 5), and Balance-ALB (Mode 6). We evaluate h2load performance under 3 Gigabit/s DDoS attacks. The results show that Round-Robin (Mode 0) can be served over HTTP/2 as a normal service during a DDoS attack. In addition, Balance-ALB (Mode 6) can be serviced like Mode 0, but this mode has lower performance compared to normal service. However, Mode 2 and Mode 5 can't be served while DDoS attack sizes exceed 20 KB. This paper can conclude that only round-robin mode can protect against DDoS attacks. Although HTTP/2 can protect against DDoS attacks, it can also work after link aggregation © 2024 IEEE.

关键词： HTTP

来源：评论

学校读者我要写书评

暂无评论

On Bounding the Behavior of a Neuron 36th

On Bounding the Behavior of a Neuron

引用

36th International Florida Artificial Intelligence Research Society Conference, FLAIRS-36 2023

作者： Borowski, Richard Choi, Arthur Computer Science Department College of Computing and Software Engineering Kennesaw State University United States

A neuron with binary inputs and a binary output represents a Boolean function. Our goal is to extract this Boolean function into a tractable representation that will facilitate the explanation and formal verification of a neuron's behavior. Unfortunately, extracting a neuron's Boolean function is in general an NP-hard problem. However, it was recently shown that prime im-plicants of this Boolean function can be enumerated efficiently, with only polynomial time delay. Building on this result, we propose a best-first search algorithm that is able to incrementally tighten inner and outer bounds of a neuron's Boolean function. These bounds correspond to truncated prime-implicant covers of the Boolean function. We provide two case studies that highlight our ability to bound the behavior of a neuron. © 2023 by the authors. All rights reserved.

关键词： Boolean functions

来源：评论

学校读者我要写书评

暂无评论

Evolutionary Insulin Management: A Digital Health Solutions for Type 1 Diabetes Support 25

Evolutionary Insulin Management: A Digital Health Solutions ...

引用

25th International Arab Conference on Information Technology, ACIT 2024

作者： Aljawawdeh, Hamzeh Salman, Randa Alzaben, Rasha Dughmosh, Ragae Khouj, Mohammed Zarqa University Department of Software Engineering Zarqa Jordan Yarmouk University Department of Computer Engineering Irbid Jordan Hamad Medical Corporation Medicine Department Doha Qatar University of Business and Technology Department of Electrical Engineering Jeddah21432 Saudi Arabia

ISBN: (纸本)9798331540012

Type 1 diabetes is characterised by a deficiency of pancreatic beta cells, resulting in hyperglycemia and insulin deficiency. It usually develops during childhood or adolescence, but symptoms can appear much later. The specific origin of type 1 diabetes is still unknown, but T cell-mediated beta cell death is thought to be responsible for the monogenesis of the disease. New treatments such as insulin pumps, continuous glucose monitoring and hybrid closed-loop systems are being developed. Tight glycemic control has reduced the incidence of microvascular and macrovascular disorders. However, most T1DM patients continue to experience these problems and significant research efforts are needed to improve the quality of life and prognosis of T1DM patients through early diagnosis, prevention of beta cell depletion and better treatment options. Exogenous insulin therapy is needed in this disease to avoid the devastating consequences of hyperglycemia. Psychological issues and comorbidity management are elements of good T1D care, proper nutrition, diabetes treatment, and comprehensive diabetes education. Additionally, several challenges must be overcome at different stages of life. Managing T1D with limited resources Healthcare professionals deal with various difficulties at work. Nonetheless, an alphabetical approach can achieve the best outcome and enable comprehensive T1D care in a scenario where resources are scarce. In the management of children with type 1 diabetes mellitus (T1DM), effective glycemic control is crucial since it is associated with improved short- and long-term health outcomes. However, poor glycemic control lowers quality of life and increases the risk of acute and chronic issues. © 2024 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Sri Potha: A Voice-Based Sinhala Document Maker

Sri Potha: A Voice-Based Sinhala Document Maker

引用

2023 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2023

作者： De Silva, D.I. Piyadali, D.S.P. Dilrukshi, M.G.S. Sri Lanka Institute of Information Technology Department of Computer Science and Software Engineering Sri Lanka

ISBN: (纸本)9798350319200

This study discusses the development of a web application aimed at facilitating Sinhala document creation, with a specific emphasis on Sinhala voice-to-text conversion and the handling of Sinhala commands through vocal input. Leveraging machine learning techniques, including convolutional neural networks and Natural Language Processing, the application's core features were established. Extensive research was conducted to tailor the application's content to the needs of its primary users, ensuring maximum effectiveness. The user-friendly interfaces of the web application are designed for clarity, simplicity, and consistency. The primary objective of this research is to comprehensively analyze the implementation of Sinhala voice-to-text conversion and Sinhala command handling systems. These systems are primarily designed to benefit diverse users, including journalists, content writers, and differently abled individuals with verbal abilities, by enhancing the efficiency of creating Sinhala documents. Through a detailed exploration of the research methodology, this study offers insight into the development process of the web-based system. The outcomes of the linguistic model training, presented within the study, reveal achievements and advancements that address the limitations inherent in existing solutions. Key findings from this research demonstrate the successful functionality of the Sinhala voice-to-text converter and the efficacy of the Sinhala command handler. The voice-to-text conversion system achieved an impressive accuracy rate of over 80%, while the Sinhala command handler exhibited an accuracy of approximately 80%. Moreover, this research envisions potential applications that extend beyond document creation. The technology showcased in the web application holds promise for broader language-based applications, impacting education, accessibility, and communication across the native Sinhala-speaking community in Sri Lanka. In summary, this research showcases th

关键词： convolutional neural network machine learning natural language processing

来源：评论

学校读者我要写书评

暂无评论

Automated Aortic Valve Calcific Area Segmentation in Echocardiography Images Using Fully Convolutional Neural Networks 47

Automated Aortic Valve Calcific Area Segmentation in Echocar...

引用

47th International Conference on Telecommunications and Signal Processing, TSP 2024

作者： Cakir, Mervenur Ekinci, Murat Kablan, Elif Baykal Sahin, Mursel Karadeniz Technical University Department of Software Engineering Trabzon Turkey Karadeniz Technical University Department of Computer Engineering Trabzon Turkey Karadeniz Technical University Faculty of Medicine Department of Cardiology Trabzon Turkey

ISBN: (纸本)9798350365597

Aortic valve calcium scoring is extensively utilized for diagnosing, treating, monitoring, and assessing the risk of aortic stenosis and coronary artery disease. The gold standard method for determining aortic valve calcium score is computed tomography (CT). However, CT is expensive and exposes patients to radiation. Alternatively, echocardiography, which is a cheaper and radiation-free method, may not provide an objective result due to inter- and intra-observer variability among cardiologists. Therefore, automated measurement of aortic valve calcium score is essential. In this paper, a fully convolutional neural network approach is proposed for the segmentation of aortic valve calcified regions from echocardiography images, as the first step in predicting aortic valve calcium score. This aims to achieve a more objective, faster, easier, less costly, and radiation-free automatic calculation of aortic valve calcium score. The proposed method achieves a precision of 86.3%, recall of 80.27%, a dice coefficient of 76.27%, and a Jaccard index of 61.9% on the new dataset including 118 echocardiography images across 82 individual patients. These results show the feasibility and potential efficacy of fully automating aortic valve calcium scoring from echocardiography images. © 2024 IEEE.

关键词： Calcium

来源：评论

学校读者我要写书评

暂无评论

A Review of Multi-omics and Machine Learning: Advancements and Applications in Cancer Research 3

A Review of Multi-omics and Machine Learning: Advancements a...

引用

3rd International Mobile, Intelligent, and Ubiquitous Computing Conference, MIUCC 2023

作者： Ayman, Hazem Bassem, Madonna Madkour, Malak Fawzi, Mariam Amr, Nada Yousry, Youmna Elgayar, Ziad Abdelraouf, Ashraf Ain Shams University Faculty of Engineering Department of Computer Engineering and Software Systems Cairo Egypt Misr International University Faculty of Computer Science Cairo Egypt

ISBN: (纸本)9798350306231

This comprehensive review starts with diving into the progress and real-world applications of combining multi-omics data analysis with machine learning techniques in cancer research. Multi-omics involves examining various biological data types like genomics, transcriptomics, proteomics, and metabolomics together to enhance our understanding of complex biological systems. By merging machine learning with multiomics data, we highlight the advantages for cancer studies, the deeper insights they yield and increased performance and results. Furthermore, we explore existing literature that showcases the integration of multi-omics and machine learning in cancer research. As part of our study, we conduct an experiment utilizing a multiomics dataset to predict the survival of breast cancer patients. We compare three distinct machine learning methods-ensemble, DeepProg, and DCAP-for survival prediction and conclude that despite the ensemble method that increased the prediction results of DeepProg over DCAP in multi-model setting, but the primitive capacity for DCAP is better in single model setting and achieves higher accuracy than DeepProg with noticeable margin 0.628 to 0.57 on C-Index metric, which strongly recommends using Denoising Autoencoder as the base for dimensionality reduction over the vanilla Autoencoder. Another empirical results conclude that using gaussian mixture model with diagonal covariance matrix for Clustering, which is used in DeepProg, might hinder the process for identifying reasonable clusters due to the assumption of no or zero correlation between different features which might not hold true in our problem. © 2023 IEEE.

关键词： Gaussian distribution

来源：评论

学校读者我要写书评

暂无评论

Communication and Computation O-RAN Resource Slicing for URLLC Services Using Deep Reinforcement Learning

IEEE Communications Standards Magazine

引用

IEEE Communications Standards Magazine 2023年第1期7卷 66-73页

作者： Filali, Abderrahime Nour, Boubakr Cherkaoui, Soumaya Kobbane, Abdellatif Department of Computer and Software Engineering Polytechnique Montreal Canada Univeriste de Sherbrooke Canada Concordia University Canada UM5R Ensias Morocco

The evolution of the future beyond-5G/6G networks toward a service-aware network is based on network slicing technology. With network slicing, communication service providers seek to meet all the requirements imposed by the verticals, including ultra-reliable low-latency communication (URLLC) services. In addition, the open radio access network (O-RAN) architecture paves the way for flexible sharing of network resources by introducing more programmability into the RAN. RAN slicing is an essential part of end-to-end network slicing since it ensures efficient sharing of communication and computation resources. However, due to the stringent requirements of URLLC services and the dynamics of the RAN environment, RAN slicing is challenging. In this article, we propose a two-level RAN slicing approach based on the O-RAN architecture to allocate the communication and computation RAN resources among URLLC end devices. For each RAN slicing level, we model the resource slicing problem as a single-agent Markov decision process and design a deep reinforcement learning algorithm to solve it. Simulation results demonstrate the efficiency of the proposed approach in meeting the desired quality of service requirements. © 2017 IEEE.

关键词： Radio access networks

来源：评论

学校读者我要写书评

暂无评论

Profiling Casualty Severity Levels of Road Accident Using Weighted Majority Voting

引用

computers, Materials & Continua 2022年第6期71卷 4609-4626页

作者： Saba Awan Zahid Mehmood Hassan Nazeer Chaudhry Usman Tariq Amjad Rehman Tanzila Saba Muhammad Rashid Department of Software Engineering University of Engineering and TechnologyTaxila47050Pakistan Department of Computer Engineering University of Engineering and TechnologyTaxila47050Pakistan Department of Electronics Information and BioengineeringPolitecnico di MilanoMilano20122Italy College of Computer Engineering and Sciences Prince Sattam bin Abdulaziz UniversityAl-Kharj11942Saudi Arabia College of Computer and Information Sciences Prince Sultan UniversityRiyadhSaudi Arabia Department of Computer Engineering Umm Al-Qura UniversityMakkah21421Saudi Arabia

To determine the individual circumstances that account for a road traffic accident,it is crucial to consider the unplanned connections amongst various factors related to a crash that results in high casualty *** of the road accident data concentrated mainly on categorizing accidents into different types using individually built classification methods which limit the prediction accuracy and fitness of the *** this article,we proposed a multi-model hybrid framework of the weighted majority voting(WMV)scheme with parallel structure,which is designed by integrating individually implemented multinomial logistic regression(MLR)and multilayer perceptron(MLP)classifiers using three different accident datasets i.e.,IRTAD,NCDB,and *** proposed WMV hybrid scheme overtook individual classifiers in terms of modern evaluation measures like ROC,RMSE,Kappa rate,classification accuracy,and performs better than state-of-theart approaches for the prediction of casualty severity ***,the proposed WMV hybrid scheme adds up to accident severity analysis through knowledge representation by revealing the role of different accident-related factors which expand the risk of casualty in a road *** aspects related to casualty severity recognized by the proposed WMV hybrid approach can surely support the traffic enforcement agencies to develop better road safety plans and ultimately save lives.

关键词： Prediction hybrid framework severity class casualty

来源：评论

学校读者我要写书评

暂无评论

The Effectiveness of Advance Deep Learning Architectures for Classification of Stress using Raw EEG Data

The Effectiveness of Advance Deep Learning Architectures for...

引用

2024 Asia Pacific Conference on Innovation in Technology, APCIT 2024

作者： Tibrewal, Ayush Shikha Sethia, Divyashikha Delhi Technological University Department of Applied Physics Delhi India Delhi Technological University Science and Engineering Department of Computer Delhi India Delhi Technological University Department of Software Engineering Delhi India

ISBN: (纸本)9798350361537

Understanding and detecting stress in today's fast-paced world remains challenging. Previous research has explored various physiological signals, including EEG (Electroencephalography), to quantify stress. However, accurately assessing stress poses a significant challenge, particularly in preprocessing and feature extraction of EEG signals. Hence, this paper employs advanced deep learning algorithms, including ResNet18-1D, DenseNet-1D, and VGG16-1D, on raw EEG signals for automated stress classification. The study utilizes the publicly available EEG-based SAM40 dataset for stress classification and further validates the results on the EEG-based MAT dataset. The results demonstrate the efficacy of the ResNet18-1D model by achieving the highest of 98.73% and 97.24% on the SAM40 and MAT datasets, respectively. The findings pave the way for further exploration and utilization of raw EEG signals across diverse research domains beyond stress classification. © 2024 IEEE.

关键词： Electroencephalography

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：