检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wang, Yunuo Yang, Ningning Li, Jialin College of Life Science and Technology Huazhong University of Science and Technology Wuhan China Department of Computer Science Data Science and Engineering New York University Shanghai Shanghai China School of Life Sciences and Biotechnology Shanghai Jiao Tong University Shanghai China

Generative Adversarial Networks (GANs) have surfaced as a revolutionary element within the domain of low-dose computed tomography (LDCT) imaging, providing an advanced resolution to the enduring issue of reconciling radiation exposure with image quality. This comprehensive review synthesizes the rapid advancements in GAN-based LDCT denoising techniques, examining the evolution from foundational architectures to state-of-the-art models incorporating advanced features such as anatomical priors, perceptual loss functions, and innovative regularization strategies. We critically analyze various GAN architectures, including conditional GANs (cGANs), CycleGANs, and Super-Resolution GANs (SRGANs), elucidating their unique strengths and limitations in the context of LDCT denoising. The evaluation provides both qualitative and quantitative results related to the improvements in performance in benchmark and clinical datasets with metrics such as PSNR, SSIM, and LPIPS. After highlighting the positive results, we discuss some of the challenges preventing a wider clinical use, including the interpretability of the images generated by GANs, synthetic artifacts, and the need for clinically relevant metrics. The review concludes by highlighting the essential significance of GAN-based methodologies in the progression of precision medicine via tailored LDCT denoising models, underlining the transformative possibilities presented by artificial intelligence within contemporary radiological practice. © 2024, CC BY-SA.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

A Novel approach for data mining Classification using J48DT Classifier for Intrusion Detection System 11

A Novel approach for Data mining Classification using J48DT ...

引用

11th IEEE International Conference on Communication Systems and Network Technologies, CSNT 2022

作者： Manimegalai, T. Ravishankar, T. Nadana Kannagi, L. Kannan, K. Anitha, G. School of Computer Science and Engineering Vellore Institute of Technology Chennai India Department of Data Science and Business System Srm Institute of Science and Technology Chennai Kattankulathur India Department of Computer and Communication Engineering Sri Sairam Institute of Technology Chennai India Department of Ece R.M.K. College of Engineering and Technology Puduvoyal Chennai India Department of Ece Saveetha School of Engineering Simats Chennai India

ISBN: (纸本)9781665480383

Because of the growing quantity of network traffic that is being transmitted across the network in today's modern era, network attack detection has become a necessity. The technique of data mining is extremely important in the hunt for network assaults and anomalies. These strategies aid in the selection and refinement of meaningful and relevant knowledge from enormous data sets, which is beneficial in many situations. The process of data mining aids in the categorisation of important information for Intrusion Detection Systems (IDS). The IDS provides warnings for the network traffic informing it of any foreign intrusions into the network. We show how different data mining methods may be used for intrusion detection to offer a safe network environment in the research article. The J48DT Classifier is utilised for classification in this case. Comparisons based on characteristics such as data accuracy, the discovery of valuable patterns, and the extraction of useful information between approaches are made. © 2022 IEEE.

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

Active Advantage-Aligned Online Reinforcement Learning with Offline data

arXiv

引用

arXiv 2025年

作者： Liu, Xuefeng Le, Hung T.C. Chen, Siyu Stevens, Rick Yang, Zhuoran Walter, Matthew R. Chen, Yuxin Department of Computer Science University of Chicago ChicagoIL United States Department of Statistics and Data Science Yale University New HavenCT United States Toyota Technological Institute at Chicago ChicagoIL United States

Online reinforcement learning (RL) enhances policies through direct interactions with the environment, but faces challenges related to sample efficiency. In contrast, offline RL leverages extensive pre-collected data to learn policies, but often produces suboptimal results due to limited data coverage. Recent efforts have sought to integrate offline and online RL in order to harness the advantages of both approaches. However, effectively combining online and offline RL remains challenging due to issues that include catastrophic forgetting, lack of robustness and sample efficiency. In an effort to address these challenges, we introduce A3RL, a novel method that actively selects data from combined online and offline sources to optimize policy improvement. We provide theoretical guarantee that validates the effectiveness our active sampling strategy and conduct thorough empirical experiments showing that our method outperforms existing state-of-the-art online RL techniques that utilize offline data. Our code will be publicly available at: https://***/xuefeng-cs/A3RL. Copyright © 2025, The Authors. All rights reserved.

关键词： Active learning

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Evolutionary Algorithm for Optimized Real-Time Scheduling in Cloud Computing

An Enhanced Evolutionary Algorithm for Optimized Real-Time S...

引用

Computational Intelligence Approaches and Applications (ICCIAA), International Conference on

作者： Asia Bataineh Hamzah Alqudah Mohammad A. Alsharaiah Waleed Awwad Computer Science Department Faculty of Information Technology University of Petra (UoP) Amman Jordan Department of Data Science and Artificial Intelligence Faculty of Information Technology American University of Madaba (AUM) Amman Jordan Information Technology Department King Abdullah II School for Information Technology The University of Jordan (JU) Amman Jordan Department of Computer Science College of Science Northern Border University (NBU) Arar KSA

ISBN: (数字)9798331523657

ISBN: (纸本)9798331523664

The exponential growth of cloud computing applications has posed significant challenges in achieving efficient resource management, reduced processing time, high through-put, and minimized energy consumption. Despite advancements, existing evolutionary algorithms often suffer from premature convergence, inefficient exploration-exploitation balance, and high energy overhead, particularly under heavy workloads. This paper proposes an enhanced hybrid Whale Optimization Algorithm (WOA) integrated with the Lévy Flight Mechanism for optimized resource management in cloud computing environments to address these challenges. The primary motivation for this work stems from the limitations observed in state-of-the-art algorithms. The research aims to bridge this gap by designing an evolutionary-based approach that improves resource utilization, reduces task execution delays, and optimizes energy consumption under varying workloads. The proposed model employs the WOA, enhanced with the Lévy Flight Mechanism, to ensure a more effective balance between exploration and exploitation during optimization. Extensive simulations were conducted on workloads ranging from 100 to 700 tasks, comparing the proposed method against leading algorithms. The results demonstrate that the proposed algorithm consistently outperforms the state-of-the-art techniques. Specifically, it achieves higher resource utilization (1.1 vs. 0.9), lower processing time, improved throughput (up to 500), and reduced energy consumption (50 joules compared to 70+ joules) under heavy workloads. These findings validate the enhanced WOA's effectiveness in addressing critical cloud computing optimization problems. The proposed approach provides a robust solution for dynamic resource allocation, ensuring energy efficiency and performance scalability, making it suitable for modern cloud-based environments with increasing computational demands.

关键词： Cloud computing Energy consumption Processor scheduling Scalability Evolutionary computation Throughput Whale optimization algorithms Real-time systems Resource management Optimization

来源：评论

学校读者我要写书评

暂无评论

Multi-Feature Driver Drowsiness Detection

Multi-Feature Driver Drowsiness Detection

引用

2021 International Conference on data Analytics for Business and Industry, ICDABI 2021

作者： Abdulrahman, Hajar Taha, Ranyia Al Haj, Duaa Al-Ammal, Hesham College of Science University of Bahrain Sakhir Big Data Science and Analytics Bahrain College of Information Technology University of Bahrain Sakhir Department of Computer Science Bahrain

ISBN: (纸本)9781665416566

Driver drowsiness is one of the leading causes of deadly car accidents, and therefore has become of great interest in the field of Artificial Intelligence. In this paper, convolutional neural networks (CNNs) are used to determine the state of the driver (i.e. Drowsy or not), based on 4 observed features: open eyes, closed eyes, yawn and no yawn. The data was collected from two open source website, which was severely imbalanced between the open/closed eyes images and yawn/no yawn images, the former two classes being the majority class. Augmentation techniques were applied, and three different model structures were investigated, where the best architecture yielded an accuracy rate of 95%, and a loss of 13%. The paper discusses the success and limitations of these models and possible future work on approaching this problem. © 2021 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Parallel Branch-and-Bound-Comparing Serial and OpenMP Implementations in a Color Sorting Game

Parallel Branch-and-Bound-Comparing Serial and OpenMP Implem...

引用

2021 International Conference on data Analytics for Business and Industry, ICDABI 2021

作者： Khalid, Aisha Abdulrahman, Hajar Khedr, Ahmed Alqaddoumi, Abdulla University of Bahrain Big Data Science and Analytics College of Science Sakhir Bahrain University of Bahrain College of Information Technology Department of Computer Science Sakhir Bahrain

ISBN: (纸本)9781665416566

Branch-and-Bound algorithm is very diverse in its applications, one of which is the gaming industry, especially for complicated games that need optimization in their search for a solution from a large search space. For a thorough study, the Branch-and-Bound algorithm will be implemented in a serial and parallel (using Open Multi-Processing) approach to compare time required to solve the game. Results showed that for easy levels, the serial program performed better, while for difficult levels (with more flasks and colors) the parallel program outperformed the serial program with a speed up of more than 1,000x. Future work can replicate the algorithm on different types of games and explore different approaches such as Backtracking in a parallel approach. © 2021 IEEE.

关键词： Branch and bound method

来源：评论

学校读者我要写书评

暂无评论

A Bioinformatics Analysis and Homology 3D Modeling of Dihydroxy-orotate-dehydrogenase of Plasmodium ovale

A Bioinformatics Analysis and Homology 3D Modeling of Dihydr...

引用

International Conference on Informatics, Multimedia, Cyber and Information System (ICIMCIS)

作者： Rudi Nirwantono Khoirunnisa Assidqi Alam Ahmad Hidayat Joko Pebrianto Trinugroho Bens Pardamean Biotechnology Department Faculty of Engineering Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Biotechnology Department Faculty of Engineering Food Biotechnology Research Center Bina Nusantara University Jakarta Indonesia Mathematics Departement School of Computer Science Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta Indonesia Computer Science Department BINUS Graduate Program – Master of Computer Science Program Bina Nusantara University Jakarta Indonesia

The infection of Plasmodium vivax is relatively less virulent than the deathliest Plasmodium falciparum. However, it still can lead to a fatal case and often induces recurring malaria due to dormant parasites in the liver. Thus, the research to study the drug to treat Plasmodium vivax is essential, where the enzyme dihydroorotate dehydrogenase (DHODH) has recently become a new drug target. However, the drug-enzyme interaction study has only recently been conducted in Plasmodium falciparum (pfDHODH) due to the lack of the 3D structure of enzyme DHODH from Plasmodium vivax that is crucial for the study. Therefore, this study aimed to perform the modelling study of Plasmodium vivax DHODH (PvDHODH) to create a 3D structure as a basis for drug-protein interaction study in upcoming studies. Sequence pvDHODH (Accession: SCO68359.1) was used in homology modelling using Modeller 10.2 with the crystal structure of DHODH from DHODH (Accession: 7KZY) as a template. Overall, the model generated from the homology modelling was considered a good model, and the 3D structure was close to the native state according to several parameters, including DOPE score, GA341 score, QMEAN4 value, 3D-1D score, ERRAT2 score, and ProSA score. Ramachandran plot also revealed that almost all amino acids were distributed in the desirable and allowed area (99.1%), and only 0.9% were in generously allowed regions. Superimposition of the model with the template also indicated that the model has an almost similar structure and amino acids positioning, including in the binding pocket and active site. Therefore, the model can be used for downstream analysis in drug-protein interaction studies. Nevertheless, some improvements can still be performed to upgrade the quality of the model by deleting unaligned residues on the C-terminal and realigning the residues at the doubting site.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A GoogLeNet Performance Approach for COVID-19 Detection using Chest X-ray Images

A GoogLeNet Performance Approach for COVID-19 Detection usin...

引用

International Conference on Knowledge and Smart Technology (KST)

作者： Patipan Rattanawin Tidatep Pakinsee Pokpong Songmuang Data Science and Innovation College of Interdisciplinary Studies Thammasat University Pathum Thani Thailand Department of Computer Science Faculty of Science and Technology Thammasat University Pathum Thani Thailand

Coronavirus disease (COVID-19) is a major pandemic disease that has already infected millions of people worldwide and affects many aspects, especially public health. There are many clinical techniques for the diagnosis of this disease, such as RT-PCR and CT-Scan. X-ray image is one of the important techniques for medical diagnosis and easily accessible in classifying suspected cases of COVID-19 infection. In this study, we classified COVID-19 images with four classes: COVID-19, Normal, Lung opacity and Viral pneumonia by compared three models: EfficientNetB0, MobileNet and GoogLeNet for the performance of classification using 1,000 chest X-ray images from Kaggle dataset within scenario of resource limitations. The experiment reveals that GoogLeNet shows superiority over other models that produced the highest accuracy results of 88% and F1 score of 0.88 with a total time of 1 hour and 15 minutes. Along with its confusion matrix that shows model can better classify images than other models.

关键词： COVID-19 Knowledge engineering Pandemics Pulmonary diseases Lung Medical diagnosis Public healthcare

来源：评论

学校读者我要写书评

暂无评论

Design and Development of Stress Monitoring System in Use Case: Twitter #Dek65

Design and Development of Stress Monitoring System in Use Ca...

引用

International Conference on Knowledge and Smart Technology (KST)

作者： Nessara Sukumthammarat Sasima Nimma Pokpong Songmuang Data Science and Innovation College of Interdisciplinary Studies Thammasat University Pathum Thani Thailand Department of Computer Science Faculty of Science and Technology Thammasat University Pathum Thani Thailand

Twitter is a social media platform where users can post, make a conversation, comments, and share experiences that express their emotions and sentiments. Our objective is to monitor and analyze the #Dek65 hashtag’s Twitter messages. We take 166,110 Twitter messages on the #Dek65 hashtag from August 2021 to July 2022 and bring them to analyze attitudes, thoughts, emotions, and stress during the preparation for university entrance exams and the Thai education system. We designed and developed a system by creating a model that can sentiment message, a model for cluster topics from the negative message then represent sentiment messages in a way that is simple to understand through visualization. We do this to make stakeholders can monitor and aware of the problems in the Thai education system.

关键词： Visualization Emotion recognition Social networking (online) Blogs Anxiety disorders Education Stakeholders

来源：评论

学校读者我要写书评

暂无评论

Urban Sound Classification Using DNN, CNN LSTM a Comparative Approach

Urban Sound Classification Using DNN, CNN LSTM a Comparative...

引用

2021 International Conference on Innovation and Intelligence for Informatics, Computing, and Technologies, 3ICT 2021

作者： Bubashait, Mohamed Hewahi, Nabil Big Data Science and Analytics Master Program College of Science University of Bahrain Sakhier Bahrain University of Bahrain Department of Computer Science College of IT Sakhier Bahrain

ISBN: (纸本)9781665440325

Like air pollution, sound pollution has grown to be a major concern for city residents, designers, and developers. Detecting and recognizing sound types and sources in cities and suburban areas or any environment have become a necessity for the quality of life as well as security. In recent years, researchers have explored many models using Convolutional Neural Network (CNN), Long Short Term Memory (LSTM) Neural Network, and different combinations of these techniques, which produced promising results when combined with spectrogram images, or its different variations, to classify urban sounds. This research uses the DNN model performance as a baseline to compare the CNN and LSTM models' performance for classifying urban sound using Mel scale cepstral analysis (MEL) spectrum images using an open-source library called Librosa for sound processing. Models' performance was evaluated using the UrbanSound8k dataset. The CNN model underperformed with an accuracy rate of 87.15% and f1 score of 85.63% compared to both the DNN base model and the LSTM model. In contrast, comparing the LSTM model with CNN, LSTM shows better accuracy performance on test data with 90.15%, and f1 score of 90.15%. © 2021 IEEE.

关键词： Long short-term memory

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：