检索结果-内蒙古大学图书馆

Quantum-Accelerated Hyperparameter Tuning for Dynamic NLP Models

Procedia computer science 2024年 235卷 506-519页

作者： Ravikumar S Arockia Raj Y R. Babu Vijay K R. Ramani Associate Professor Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai Tamilnadu India 600062 Assistant Professor Department of Computer Science and Engineering PSNA College of Engineering and Technology Dindigul India Assistant Professor Department of Computational Intelligence Faculty of College of Engineering and Technology SRM Institute of Science and Technology Chennai Tamilnadu India 603203 Assistant Professor Department of Computer Science and Engineering Rajalakshmi Engineering College Chennai Tamilnadu India 602117 Associate Professor Department of Computer Science and Engineering P.S.R. Engineering College Sivakasi Tamilnadu India 626140

In the rapidly evolving field of natural language processing (NLP), performance optimization of large-scale NLP models is crucial. Through the application of Quantum-Accelerated Hyperparameter Tuning (QAHT), this abstract introduces a novel approach to addressing this issue. Our proposed framework leverages quantum computing capabilities to dynamically optimize NLP model hyperparameters in real-time, catering to the ever-changing character of textual data streams. Traditional hyperparameter optimisation methods usually rely on laborious grid searches or random exploration, which may not be suitable for dynamic NLP jobs. Contrarily, QAHT uses Quantum Neural Network (QNN) architectures that have been specially designed for hyperparameter optimisation. These QNNs improve performance and efficacy by dynamically modifying and improving model configurations. This abstract discusses the key elements of the QAHT architecture, including real-time model deployment, adaptive learning, and continuous data stream processing. In addition to speeding up the hyperparameter optimisation process, QAHT makes sure that NLP models are still flexible and responsive to shifts in the sentiment of the data and its distribution. This method has applications beyond NLP since it provides a foundation for effectively optimising machine learning models in complex, real-time situations. As quantum computing develops, QAHT represents a promising future in machine learning, where quantum-enhanced capabilities satisfy the needs of contemporary data-driven applications.

关键词： Hyperparameter optimization Quantum computing Natural language processing (NLP) Quantum Neural Networks (QNN) Real-time processing Adaptive learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing Educational Video Discovery Using Advanced Latent Semantic Analysis

引用

Procedia computer science 2025年 252卷 784-795页

作者： B. Sindhu A. Bhaskar G. Yugesh S. Reshma B. Rohit Assistant Professor Department of Computer Science and Engineering Godavari Global University Rajamahendravaram -533296 Andhra Pradesh India

The rapid growth of online educational videos has created abundant learning resources, but also has challenges in discovering high-quality, relevant content. This project mainly focuses on video-based content analysis and providing high-quality learning sources for scholars, easing the best possible way to discover accurate resources. Existing Systems often rely on round-robin, random selection which fails to grasp semantics and linguistic complexity. The proposed system focuses on content-based analytics using NLP, where the extraction of transcripts from videos, perform relevance assessment which maps content to concept and extract insights using Latent Semantic Analysis which effectively captures underlying structures in the transcript. Complexity Assessment measures readability metrics using the Flesch-Kincaid Grade Level and SMOG Index. Created a web-based interface for easy access to educational videos, with built-in analytics to assess relevance. The insights are shared with scholars to refine content and better meet user requirements.

关键词： Latent Semantic Analysis NLP(Natural Language Processing) SMOG(Simple Measure of Gobbledygook) Index Topic Modelling TF-IDF (Term frequency-inverse document frequency) Content-based filtering

来源：评论

学校读者我要写书评

暂无评论

Artificial intelligence in the detection and classification of dental caries

引用

Journal of Prosthetic Dentistry 2025年第5期133卷 1326-1332页

作者： Ahmed, Walaa Magdy Azhari, Amr Ahmed Fawaz, Khaled Ahmed Ahmed, Hani M. Alsadah, Zainab M. Majumdar, Aritra Carvalho, Ricardo Marins Assistant Professor Department of Restorative Dentistry Faculty of Dentistry King Abdulaziz University Jeddah Saudi Arabia Associate Professor Department of Orthopedic Surgery Faculty of Medicine Cairo University Cairo Egypt Assistant Professor Department of Civil Engineering Faculty of Engineering King Abdulaziz University Jeddah Saudi Arabia Consultant in Restorative Dentistry Dental Department East Jeddah General Hospital Ministry of Health Jeddah Saudi Arabia Graduate student Department of Computer Science Computer Science and Applications Virginia Polytechnic Institute and State University Blacksburg VA United States Professor Department of Oral Biological and Medical Sciences Faculty of Dentistry University of British Columbia Vancouver BC Canada

Statement of problem: Automated detection of dental caries could enhance early detection, save clinician time, and enrich treatment decisions. However, a reliable system is lacking. Purpose: The purpose of this study was to train a deep learning model and to assess its ability to detect and classify dental caries. Material and methods: Bitewings radiographs with a 1876×1402-pixel resolution were collected, segmented, and anonymized with a radiographic image analysis software program and were identified and classified according to the modified King Abdulaziz University (KAU) classification for dental caries. The method was based on supervised learning algorithms trained on semantic segmentation tasks. Results: The mean score for the intersection-over-union of the model was 0.55 for proximal carious lesions on a 5-category segmentation assignment and a mean F1 score of 0.535 using 554 training samples. Conclusions: The study validated the high potential for developing an accurate caries detection model that will expedite caries identification, assess clinician decision-making, and improve the quality of patient care. © 2025 Editorial Council for The Journal of Prosthetic Dentistry

关键词：

来源：评论

学校读者我要写书评

暂无评论

Heart Rate Anomaly Detection in Healthcare Using Elliptic Envelope and Local Forest

引用

Procedia computer science 2025年 258卷 1677-1687页

作者： Chandralekha E S. Vinodhini Kandasamy V Rama P Assistant Professor Department of Computer Science and Engineering Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology Chennai Tamilnadu India-600062 Assistant Professor Department of Information Technology Velammal Engineering College Chennai Tamilnadu India-600062 Assistant Professor Panimalar Engineering College Chennai Tamilnadu India – 600123 Assistant Professor Department of Computing Technologies College of Engineering and Technology Faculty of Engineering and Technology SRM Institute of Science and Technology Chennai Tamilnadu India-603203

Medical professionals use Anomaly Detection(AD) to identify patients with potential health problems in the heart rate data (HRD), an essential metric related to cardiovascular health. In this research, the most effective AD techniques were identified in detecting unusual variations of the heart rate pattern. Synthetic HRD data with anomalous rates of 15%, 20%, and 25% was generated and standardized it using min-max scaling for consistency among the datasets. Normal HR was based on function with a mean of 70 and a standard deviation of 5 while anomalous values were generated with a mean of 120 and a standard deviation of 10. Among the evaluated models (Elliptic Envelope, one-class SVM, Local Outlier Factor, and Isolation Forest), the Elliptic Envelope (EE) and Isolation Forest (IF) models obtain the best scores in terms of recall and precision – 0.92 and 0.89, respectively. Overall, the results of this research indicate that equally robust AD models should be implemented within clinical systems and wearable devices that serve as a digital space for early cardiac risk identification, prompting adequate intervention and, as a result, health improvement.

关键词： Anomaly Detection Heart Rate Data Isolation Forest Local Outlier Factor One-Class SVM Elliptic Envelope

来源：评论

学校读者我要写书评

暂无评论

Evaluating Generative Adversarial Networks for Virtual Contrast-Enhanced Kidney Segmentation using Res-UNet in Non-Contrast CT Images

引用

Multimedia Tools and Applications 2025年第18期84卷 20121-20144页

作者： Syamala, Maganti Chandrasekaran, Raja Balamurali, R. Rani, R. Hashmi, Arshad Kiran, Ajmeera Rajaram, A. Assistant Professor Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Guntur District Andhra Pradesh Vaddeswaram522502 India Department of ECE Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science and Technology Tamil Nadu Chennai India Centre for non linear systems Chennai institute of technology Chennai India Department of ECE Vemu Institute of Technology Tirupathi-chittoor Highway P. Kothakota Chittoor District AP Pakala India King Abdulaziz University P.O. Box 344 Rabigh21911 Saudi Arabia Department of Computer Science and Engineering MLR Institute of Technology Dundigal Telangana Hyderabad500043 India Department of Electronics and Communication Engineering EGS Pillay Engineering College Nagapattinam611002 India

This study addresses the challenges of accurately segmenting kidneys in non-contrast CT images;where Convolutional neural networks (CNNs) perform best. The main goal is to investigate how well a Generative Adversarial Network (GAN) performs in producing virtual contrast-enhanced (vCECT) images from non-contrast sources. A GAN model is trained to generate simulated contrast-enhanced images that preserve pertinent information throughout translation by utilizing a dataset of 286 paired non-contrast and contrast-enhanced images. Following, 20 pairs of non-contrast and vCECT pictures are used to assess the performance of a unique segmentation model named Res-U-Net. The Res-U-Net architecture, which combines a U-Net design with residual connections, is optimized to efficiently extract detailed information from original and vCECT images. Image preprocessing methods like Hounsfield windowing with histogram equalization are used to improve the quality of the data before it is fed into the deep neural network. Dice similarity coefficient (DSC) is used to measure Res-U-Net's performance and acts as a benchmark against the most advanced kidney segmentation techniques. ResU-Net's performance is evaluated using the DSC. When compared to state-of-the-art methods for kidney segmentation recognition, the ResU-Net systems with intact connections achieved a DSC value of 0.97 percent for organs recognition and 0.83 percent for segmentation approaches. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Investigation of Security Attacks in IoMT Devices and Federated Learning as a Mitigation Strategy

引用

Procedia computer science 2025年 258卷 3426-3435页

作者： Kiruthika M T Poongodi Assistant Professor Department of Computational Intelligence School of Computing SRM Institute of Science and Technology Kattankulathur Tamilnadu 603203 India Professor Department of Computer Science and Engineering School of Engineering Dayananda Sagar University Bangalore Karnataka India

Internet of Medical Things (IoMT) is transforming the healthcare business with an increasing number of wearables, gadgets, and sensing devices. As a result, several patients are under daily monitoring for their cardio health, blood glucose level, blood pressure, etc. The data generated by these devices are very voluminous, and there is an increasing need for safeguarding these sensitive data in order to protect them from malicious attacks. This paper suggests to incorporate Federated learning model to IoMT as a mitigative measure. It is apt to induct FL into IoMT as it operates on a collaborative environment which do not allow transfer of the raw data. Patients’ data are stored at the client end itself and only model updates are shared to the server. In this way, the data and network attacks can be reduced to a great extent. This work proposes an integrated architecture where the layers of FL and IoMT can be knotted up for ensuring privacy and security. This proposed architecture model, further elaborates a distributed working environment. This paper illustrates how FL helps in securing the data produced by the different IoMT devices through the proposed architecture. The challenges involved in the deployment of FL in the framework of IoMT along with potential research directions are discussed further in this research.

关键词： : Federated Learning IoMT Security Internet of Medical Things IoMT Data Attacks Optimization

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Cerebral Splitting: Adaptation for Spine Elimination

Deep Learning Cerebral Splitting: Adaptation for Spine Elimi...

引用

Communication, Computing and Energy Efficient Technologies (I3CEET), International Conference on

作者： M. Balasubramani B. J. Job Karuna Sagar K. C. Rajheshwari E. Saranya B. Aishwarya P. Murugeswari Assistant Professor Department of Computer Science and Engineering V.S.B. Engineering College Professor Department of Computer Science and Engineering Joginpally B.R Engineering College Hyderabad India Department of Computer Science and Engineering Sona College of Technology Assistant Professor Department of Computer Science and Engineering Sri Eshwar College of Engineering Coimbatore Assistant Professor Department of Computer Science and Engineering(Cyber Security) Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Avadi Department of Electronics and Communication Engineering Theni Kammavar Sangam College of Technology

ISBN: (数字)9798331541583

ISBN: (纸本)9798331541590

The posterior fossa volume and the total intracranial volume are important factors to consider when doing a brain volumetric analysis using structural electromagnetic resonance imaging. The thorough segmentation of the whole cortex offers a method for non-invasively evaluating different brain regions. In addition, the extraction of skulls in a way that maintains anonymity is increasingly being used as a means of disseminating large amounts of data obtained from brain images. Therefore, generalizing machine learning brain regions for intracranial measurements after skull removal is an interesting task. However, data collecting is difficult due to the small number of handwritten and whole-brain atlases marked with his TICV/PFV. In this study, we use U-Net tiles to estimate intracranial volume and perform whole-brain segmentation with and without cranial characteristics simultaneously. In the absence of individually recorded total cortical volume, a transfer learning approach has been shown to estimate extra her TICV and PFV labeling during whole-brain delineation on T1-weighted MRI. More specifically, U-Net tiles are pre-trained using a big Brain COLOR atlas generated using multi-atlas segmentation, and do not include his PFV labels. The pre-trained model is improved to learn more His TICV and PFV labels using the limited Brain COLOR atlas. We also elaborate on his method for processing MR images of the brain in the absence of a skull. The results show that our method produces reasonable segmentation and size estimation results in terms of average contact distance, cubic similarity index, and total volumetric similarity for both brains within and outside the skull. is showing off. This technology is currently available as open source to the general public.

关键词： Deep learning Image segmentation Image color analysis Magnetic resonance imaging Transfer learning Pipelines Estimation Skull Resonance Telecommunication computing

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Techniques Advancements in Apple Leaf Disease Detection

引用

Procedia computer science 2024年 235卷 713-722页

作者： E Kannan Carmel Mary Belinda M J Alex David S Ruth Naveena N Almas Begum Hemalatha D Professor Assistant Professor Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai Tamilnadu India 600062 Professor Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical And Technical Sciences Chennai Tamilnadu India 602105 Assistant Professor (SG) Hindustan Institute of Technology and Science Chennai Tamilnadu India 603103 Assistant Professor Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai Tamilnadu India 600062

An exhaustive and comprehensive investigation was undertaken to address the critical issue of disease detection on apple leaves using cutting-edge deep learning techniques. The research delved into an array of diverse approaches, meticulously examining their efficacy and performance in disease detection, ultimately offering valuable insights into this vital domain. The research effort was marked by the exploration and application of a wide spectrum of deep learning models, each chosen for its distinct characteristics and potential advantages. The results of this extensive work were nothing short of remarkable. This study uses state-of-the-art deep learning techniques to present a thorough and rigorous analysis into the important problem of disease detection on apple leaves. Our research covers a wide range of approaches, all of which have been thoroughly assessed for their efficacy in the diagnosis of disease. We used a wide range of deep learning models, selected for their special qualities and possible benefits. The results of this extensive study are impressive and measurable. VGG-INCEP, the top approach, showed exceptional performance with a measured accuracy rate of 97%. The quantification of precision, recall, and F1 scores were 0.94, 0.92, and 0.92, respectively. Similarly, InceptionV3 yielded an F1 score of 0.93, precision of 0.95, and recall of 0.91, in addition to a measured accuracy of 97%. AlexNet consistently demonstrated measurable high precision (0.95) and recall (0.93), resulting in an F1 score of 0.93, despite a somewhat lower accuracy of 87%. The method's balanced performance is highlighted by these metrics. The study also evaluated the effectiveness of SVM, MobileNet, RCNN, and a recommended method. With quantifiable accuracy of 98% and quantifiable precision, recall, and F1 scores of 0.96, the suggested technique stood out. This assessment unequivocally shows that the suggested approach produces the best accuracy and overall performance and is di

关键词： Apple leaf diseases Deep learning Disease detection Precision agriculture Performance evaluation

来源：评论

学校读者我要写书评

暂无评论

Plastic Waste Management Strategies: A Sustainable Approach

Plastic Waste Management Strategies: A Sustainable Approach

引用

National Conference on Technological Advancements in Waste Management: Challenges and Opportunities, TAWMCO 2023

作者： Pandey, Srijan Singh, Sudheer Kumar Singh, Ranjana Scholar Department of Economics National P. G. College UP Lucknow India Associate Professor Department of Computer Science and Engineering Galgotias University UP Greater Noida India Assistant Professor Department of Economics National P. G. College UP Lucknow India

ISBN: (纸本)9789819760237

Plastic pollution has become a global environmental crisis, with detrimental impacts on ecosystems, economy, and human health. To address this challenge, effective plastic waste management strategies are essential. This research paper focuses on the issue of plastic waste and the management strategies to be adopted in India. The paper emphasizes the necessity of innovative waste management techniques to tackle this problem on a larger scale. Despite notable efforts to tackle the problem in India, an alarming plastic waste approximately 3.8 million tons remains uncollected or improperly disposed of each year, which contributes to pollution. Drawing on existing literature and case studies, the paper examines the environmental, economic, and social dimensions of plastic waste management. The findings underscore the importance of holistic and collaborative efforts involving government policies, industry initiatives, public awareness campaigns, and technological innovations to achieve a circular economy and mitigate the adverse effects of plastic pollution. Key strategies include the implementation of extended producer responsibility (EPR) laws, the promotion of circular economy principles, investment in recycling infrastructure, public awareness campaigns, and international cooperation. The paper offers a set of recommendations that are based on solid evidence and propose a more sustainable and environmentally friendly approach. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Circular economy

来源：评论

学校读者我要写书评

暂无评论

A novel random finite element model for holistically modeling of the frost effects on soils and cold region pavements

引用

Journal of Infrastructure Preservation and Resilience 2022年第1期3卷 1-14页

作者： Dong, Shaoyang Jiang, Yusheng Yu, Xiong Graduate Research Assistant Department of Civil and Environmental Engineering Case Western Reserve University ClevelandOH United States Opal J. and Richard A. Vanderhoof Professor and Chair Department of Civil and Environmental Engineering Professor of Electrical Engineering and Computer Science Case Western Reserve University ClevelandOH United States

This paper describes the development of a random finite element model (RFEM) that allows holistic simulation of frozen soil behaviors, including the effects of phase transition and the consequent internal stress and volume changes. The performance of the model is firstly validated with laboratory experiments. The model is implemented to simulate the effects of frost action on pavement. The coupled thermal-mechanical actions including the mechanical responses of subgrade soils subjected to freezing temperature and its effects on the pavement structure are analyzed. The results show that the frost action and expansion of ice lenses change the interaction modes between pavement layers. This leads to increase of stress and deformation in the pavement layer. Methods to mitigate the effects of frost heave are analyzed with this model. The simulation results indicate that the detrimental effects of frost heave on the pavement structure can be mitigated by increasing the thickness of base layer, use of thermal insulation layer or improve drainage in the subgrade layer. The RFEM combines the advantages of discrete element model (DEM) in holistically describing the microstructure effects and in the finite element model (FEM) in terms of computational efficiency. This allows to focus research on understanding the behaviors of individual soils phase and their interfacial interactions. © The Author(s) 2022.

关键词： Pavements

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：