检索结果-内蒙古大学图书馆

22nd International Learning and Technology Conference, L and T 2025

作者： Sanaullah Honda, Hirotada Roy, Kaushik Schneider, Axel Wasmuth, Joachim Jungeblut, Thorsten University of Applied Sciences and Arts Dept. of Mathematics and Engineering Bielefeld Germany Toyo University Dept. of Information Networking Tokyo Japan A&T State University Dept. of Computer Science NC United States

ISBN: (纸本)9798331532215

The selection of optimal neural models in Spiking Neural Networks (SNNs) traditionally depends on a trial-and-error approach, which is both time-consuming and sometimes tends to suboptimal selection of the neural model. This research study explores the integration of Automated Machine Learning (AutoML) techniques into SNNs to simplify the process of designing an SNN model by automatically selecting the most optimal neural model. For example, in traditional neural networks, AutoML proves that it has the potential to significantly reduce the time and effort required to identify the best-performing architectures, by automating the selection and updating of loss functions and hyper-parametric values. To this end, we present a stand-alone framework for applying AutoML in SNNs, highlighting how this approach can improve accuracy, efficiency, and reliability in the selection process of optimal neural models, as these neural models are the core principle of designing SNN architecture. The proposed method not only accelerates the development of SNN models but also enhances their performance by systematically identifying optimal configurations that might be overlooked through manual methods. Therefore, to validate our approach, the proposed architecture was tested using different well-known benchmarks with different sets of neurons, ranging from 100 to 3000 neurons, and demonstrated state-of-the-art results in image classification tasks. © 2025 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

A Web-Based Interface that Leverages Machine Learning to Assess an Individual’s Vulnerability to Brain Stroke

引用

IEEE Access 2025年 13卷 87950-87964页

作者： Bhandari, Divyansh Agarwal, Arnav Reena Roy, R. Priyatharshini, Rajaram Cristian, Rodriguez Rivero Vellore Institute of Technology School of Electronics Engineering Tamil Nadu Chennai India Vellore Institute of Technology School of Computer Science and Engineering Tamil Nadu Chennai India Cardiff Metropolitan University Cardiff School of Technologies Centre for Engineering Research in Intelligent Sensors and SystemsComputer Science Wales United Kingdom Cardiff Metropolitan University Academic Associate Cardiff School of Technologies Cardiff United Kingdom Universitat Politècnica de Catalunya Industrial and ICT Systems Engineering Dept. of Mining Spain

Cerebral stroke is a major global health issue, contributing to high mortality and long-term disability. Early identification of individuals at high risk of stroke can significantly improve preventive care outcomes. We present a web-based stroke risk assessment tool that uniquely combines an accessible user interface with robust machine learning modeling. The proposed platform leverages a novel combination of SMOTE oversampling and logistic regression to address class imbalance in patient health records, improving the detection of stroke risk factors over existing methods. We compare a range of algorithms – including traditional classifiers and deep learning models – and report comprehensive performance metrics (accuracy, precision, recall, F1-score, and AUC-ROC) for each. Our best model (logistic regression with SMOTE and standard scaling) achieves 93.2% accuracy with a substantially higher F1-score for the stroke-positive class than other models, indicating improved sensitivity to stroke cases. To bridge the gap between complex predictive models and end-users, we deploy this model in an intuitive web interface (built with Streamlit) that non-technical individuals and healthcare providers can easily use. This interface requires no specialized knowledge, preserves user privacy by avoiding any data storage, and provides clear explanations of results. By offering a practical and transparent tool for stroke risk screening, our work advances health informatics with an emphasis on accessibility, interpretability, and early intervention. Potential applications range from personal health self-assessment to integration in clinical workflows for preventive care, ultimately aiming to improve public health outcomes through early detection and intervention in stroke. © 2013 IEEE.

关键词： Health insurance

来源：评论

学校读者我要写书评

暂无评论

Enhancing 3D Modeling Quality Through Orthorectified Image Accumulation

Enhancing 3D Modeling Quality Through Orthorectified Image A...

引用

2025 International Conference on Electronics, Information, and Communication, ICEIC 2025

作者： Kwag, Cheolsu Kim, Euntaek Lee, Yongsuk Jung, Seungmin Hwang, Sung Soo Ryu, Hwangryol Moon, Youngmin Handong Global University Dept. of Computer Science and Electrical Engineering Pohang Korea Republic of School of Computer Science and Electrical Engineering Handong Global University Pohang Korea Republic of Research Institute of Industrial Science and Technology Smart Engineering Research Laboratory Pohang Korea Republic of

ISBN: (纸本)9798331510756

This study introduces a novel methodology designed to facilitate the capture of comprehensive image datasets, crucial for accurate 3D modeling of expansive indoor spaces. Leveraging orthophotos generated from panoramic camera data and visualizing image coverage density through heatmaps, our approach effectively identifies under-captured and unvisited areas. This enables targeted image acquisition, minimizes redundant reshoots, and enhances the overall quality and fidelity of 3D models. © 2025 IEEE.

关键词： Digital elevation model

来源：评论

学校读者我要写书评

暂无评论

Defense and Recovery Strategies for Flash-Based Storage Under Ransomware Attacks: A Survey

Defense and Recovery Strategies for Flash-Based Storage Unde...

引用

2025 International Conference on Electronics, Information, and Communication, ICEIC 2025

作者： Yin, Junhong Kang, Kyungtae Major in. Bio Artificial Intelligence Hanyang University Dept. Computer Science and Engineering Ansan Korea Republic of Hanyang University Dept. Artificial Intelligence Ansan Korea Republic of

ISBN: (纸本)9798331510756

As ransomware evolves, traditional OS-based de-tection mechanisms face growing challenges, particularly from advanced attacks exploiting system vulnerabilities and escalating privileges. The continuous evolution of ransomware attacks has driven researchers to explore storage-level defense strategies, focusing on the potential of Solid-State Drives (SSDs) to enhance data protection and recovery. This paper presents a compre-hensive survey of the current state and future developments in SSD firmware-level ransomware detection and recovery, based on literature published between 2015 and 2024. Firmware-level I/O monitoring helps detect ransomware by identifying the typical 'read-encrypt-overwrite' pattern. This low-level access offers an advantage in detecting ransomware activity. However, conventional detection methods are often prone to false pos-itives. To enhance detection accuracy, some systems integrate machine learning models within the SSD controllers, analyzing I/O request patterns and entropy values of written data to assess potential threats. The Garbage Collection(Gc)mechanism plays a crucial role in data recovery by retaining older versions of data until reclaimed, which can aid in restoring encrypted content. Similarly, backing up legitimate I/O operations incurs performance overhead. To address this, some systems utilize techniques such as neural networks and remote backups to reduce unnecessary performance degradation. Future research should prioritize developing lightweight solutions tailored for SSDs with limited computational resources. Emphasis should be placed on enhancing the timeliness of real-time detection to effectively respond to the rapidly evolving nature of ransomware attacks. © 2025 IEEE.

关键词： Firmware

来源：评论

学校读者我要写书评

暂无评论

Enhancing Livestock Supply Chains with Blockchain Traceability from Source to Market: A Survey 5

Enhancing Livestock Supply Chains with Blockchain Traceabili...

引用

5th International Conference on Emerging systems and Intelligent Computing, ESIC 2025

作者： Dahariya, Tukendra Kumar Sharma, Suraj Ray, Niranjan Kumar Guru Ghasidas Vishwavidyalaya Dept. Computer Science & Engineering Chhattisgarh Bilaspur India Kalinga Institute of Industrial Technology School of Computer Engineering Odisha Bhubaneswar India

ISBN: (纸本)9798331522100

This document provides an in-dept. analysis of how blockchain technology can greatly improve transparency, traceability, and accountability in fish and livestock supply chains, presenting the potential to transform this industry. Utilizing blockchain's decentralized framework, which guarantees a secure and unchangeable record of transactions, problems associated with inefficiencies, fraud, and data manipulation in conventional supply chains can be efficiently resolved. This survey analyzes the present condition of blockchain applications in fish and livestock management, emphasizing their advantages and drawbacks. Furthermore, it highlights significant technical and regulatory challenges that impede wider acceptance, while recognizing potential future research avenues that could enhance processes from the production stage to the retail stage. The results intend to assist stakeholders in developing supply chains that are more sustainable, efficient, and transparent, fostering trust and complying with regulatory requirements. © 2025 IEEE.

关键词： Records management

来源：评论

学校读者我要写书评

暂无评论

Advanced Tool Learning and Selection System (ATLASS): A Closed-Loop Framework Using LLM

arXiv

引用

arXiv 2025年

作者： Haque, Mohd Ariful Williams, Justin Siddique, Sunzida Islam, Md. Hujaifa Ali, Hasmot Gupta, Kishor Datta George, Roy Dept. of Cyber Physical Systems Clark Atlanta University United States Dept. of Computer Science and Engineering Daffodil International University Bangladesh Dept. of Computer Science and Engineering Ahsanullah University of Science and Technology Bangladesh

The combination of LLM agents with external tools enables models to solve complex tasks beyond their knowledge base. Human-designed tools are inflexible and restricted to solutions within the scope of pre-existing tools created by experts. To address this problem, we propose ATLASS, an advanced tool learning and selection system designed as a closed-loop framework. It enables the LLM to solve problems by dynamically generating external tools on demand. In this framework, agents play a crucial role in orchestrating tool selection, execution, and refinement, ensuring adaptive problem-solving capabilities. The operation of ATLASS follows three phases: The first phase, Understanding Tool Requirements, involves the Agents determining whether tools are required and specifying their functionality;the second phase, Tool Retrieval/Generation, involves the Agents retrieving or generating tools based on their availability;and the third phase, Task Solving, involves combining all the component tools necessary to complete the initial task. The Tool Dataset stores the generated tools, ensuring reusability and minimizing inference cost. Current LLM-based tool generation systems have difficulty creating complex tools that need APIs or external packages. In ATLASS, we solve the problem by automatically setting up the environment, fetching relevant API documentation online, and using a Python interpreter to create a reliable, versatile tool that works in a wider range of situations. OpenAI GPT-4.0 is used as the LLM agent, and safety and ethical concerns are handled through human feedback before executing generated code. By addressing the limitations of predefined toolsets and enhancing adaptability, ATLASS serves as a real-world solution that empowers users with dynamically generated tools for complex problem-solving. © 2025, CC BY-SA.

关键词： Reusability

来源：评论

学校读者我要写书评

暂无评论

Evaluation of Machine Learning Models in Student Academic Performance Prediction 5

Evaluation of Machine Learning Models in Student Academic Pe...

引用

5th International Conference on Advanced Research in Computing, ICARC 2025

作者： Sandeepa, A.G.R. Mohottala, Sanka General Sir John Kotelawala Defence University Dept. of Computer Engineering Ratmalana Sri Lanka Sri Lanka Institute of Information Technology Dept. of Computer Science Malabe Sri Lanka

ISBN: (纸本)9798331530983

This research investigates the use of machine learning methods to forecast students' academic performance in a school setting. Students' data with behavioral, academic, and demographic details were used in implementations with standard classical machine learning models including multi-layer perceptron classifier (MLPC). MLPC obtained 86.46% maximum accuracy for test set across all implementations while for train set, it was 99.45%. Under 10-fold cross validation, MLPC obtained 79.58% average accuracy for test set while for train set, it was 99.65%. MLP's better performance over other machine learning models strongly suggest the potential use of neural networks as data-efficient models. Feature selection approach played a crucial role in improving the performance and multiple evaluation approaches were used in order to compare with existing literature. Explainable machine learning methods were utilized to demystify the black box models and to validate the feature selection approach. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Skin Disease Detection and Classification of Actinic Keratosis and Psoriasis Utilizing Deep Transfer Learning

arXiv

引用

arXiv 2025年

作者： Ahmmed, Fahud Raihan, Md. Zaheer Nahar, Kamnur Asadujjaman, D.M. Rahman, Md. Mahfujur Tamim, Abdullah Dept. of Computer Science & Engineering Varendra University Rajshahi Bangladesh Dept. of Computer Science & Engineering Khulna University of Engineering & Technology Bangladesh Dept. of Computer Science & Engineering Rajshahi University of Engineering & Technology Bangladesh Dept. of Computer Science & Engineering Rajshahi University Bangladesh

Skin diseases can arise from infections, allergies, genetic factors, autoimmune disorders, hormonal imbalances, or environmental triggers such as sun damage and pollution. Skin diseases such as Actinic Keratosis and Psoriasis can be fatal. These are treatable if identified early. However, its diagnostic methods are expensive and not widely accessible. In this study, a novel and efficient method for diagnosing skin diseases using deep learning techniques has been proposed. This approach employs a modified VGG16 Convolutional Neural Network (CNN) model. This model includes several convolutional layers. The VGG16 model has been employed using ImageNet weights and modified top layers. The top layer is modified by fully connected layers and a final softmax activation layer to obtain the result. The dataset analyzed is publicly available and titled "Skin Disease Dataset". The VGG16 architecture does not include augmentation by default;data augmentation is typically performed through rotation, shifting, and zooming during preprocessing prior to model training. The proposed methodology achieved 90.67% accuracy using the modified VGG16 model, demonstrating reliability in classifying skin diseases. The modified pre-trained model showed promising results, increasing its potential for real-world *** Codes 68T07 © 2025, CC BY.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Contention-Free Configured Grant Scheduling for 5G URLLC Traffic 23

Contention-Free Configured Grant Scheduling for 5G URLLC Tra...

引用

Proceedings of the 60th Annual ACM/IEEE Design Automation Conference

作者： Tianyu Zhang Xiaobo Sharon Hu Song Han Dept. of Computer Science and Engineering University of Connecticut Dept. of Computer Science and Engineering University of Notre Dame

ISBN: (纸本)9798350323481

5G networks are being designed to support ultra reliable and low latency communication (URLLC) services in many real-time industrial applications. The conventional grant-based dynamic scheduling can hardly fulfill the URLLC requirements due to the non-negligible transmission delays introduced during the spectrum resource grant process. To address this problem, 5G defines a grant-free transmission scheme, namely configured grant (CG) scheduling, for uplink (UL) traffic to pre-allocate spectrum resource to user equipments (UEs). This paper studies CG scheduling for periodic URLLC traffic with real-time and collision-free guarantees. An exact solution based on Satisfiability Modulo Theory (SMT) is first proposed to generate a feasible CG configuration for a given traffic set. To enhance scalability, we further develop an efficient graph-based heuristic consisting of an offset selection method and a multicoloring algorithm for spectrum resource allocation. Extensive experiments are conducted using 3GPP industrial use cases to show that both approaches can satisfy the real-time and collision-free requirements, and the heuristic can achieve comparable schedulability ratio with the SMT-based approach but require significantly lower running time.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Point Positional Insertion Tuning for Small Object Detection

Multi-Point Positional Insertion Tuning for Small Object Det...

引用

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

作者： Kanoko Goto Takumi Karasawa Takumi Hirose Rei Kawakami Nakamasa Inoue Dept. of Computer Science Institute of Science Tokyo Tokyo Japan Dept. of Systems and Control Engineering Institute of Science Tokyo Tokyo Japan

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Small object detection aims to localize and classify small objects within images. With recent advances in large-scale vision-language pretraining, finetuning pretrained object detection models has emerged as a promising approach. However, finetuning large models is computationally and memory expensive. To address this issue, this paper introduces multi-point positional insertion (MPI) tuning, a parameter-efficient finetuning (PEFT) method for small object detection. Specifically, MPI incorporates multiple positional embeddings into a frozen pretrained model, enabling the efficient detection of small objects by providing precise positional information to latent features. Through experiments, we demonstrated the effectiveness of the proposed method on the SODA-D dataset. MPI performed comparably to conventional PEFT methods, including CoOp and VPT, while significantly reducing the number of parameters that need to be tuned.

关键词： Computational modeling Object detection Signal processing Feature extraction Encoding Acoustics Speech processing Tuning

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：