检索结果-内蒙古大学图书馆

An optimized scheduling algorithm for prioritized tasks with shared resources in cloud edge computing

Expert Systems with Applications 2025年 293卷

作者： V. Arulkumar K. Kishore Raju Rambabu Pemula M.S. Antony Vigil Department of Information Technology Easwari Engineering College Chennai Ramapuram Tamil Nadu 600089 India Department of Information Technology S.R.K.R Engineering College (A) Bhimavaram Andhra Pradesh 534204 India Department of Information Technology Vidya Jyothi Institute of Technology Hyderabad Telangana 500075 India Department of Computer Science and Engineering SRM Institute of Science and Technology Chennai Ramapuram Tamil Nadu 600089 India

Cloud computing plays a crucial role in managing the vast amounts of data generated by smart devices. However, existing task scheduling models face significant challenges in optimizing completion time, energy consumption, and rent costs. Current models often fail to balance these factors effectively, leading to inefficiencies in computational resource allocation and increased operational costs. Additionally, cloud-edge computing introduces further complexity, including time-varying channels, dynamic computational resources, and task interdependencies, all of which exacerbate the difficulty of optimizing these key metrics. To address these challenges, this work proposed a novel Hybridization of the Bald Eagle Search and Border Collie Optimization (HBES-BCO) algorithm. The novelty of this approach lies in its unique integration of two advanced optimization techniques such as Bald Eagle Search (BES) and Border Collie Optimization (BCO) to create a hybrid algorithm that can adapt to the constantly changing cloud-edge environment. By combining the exploration power of BES with the exploitation strength of BCO, the proposed HBES-BCO algorithm can effectively balance multi-objective optimization, achieving superior task scheduling. The HBES-BCO algorithm is designed to minimize the critical metrics of completion time, energy consumption, and rent costs, all while effectively managing task interdependencies. Unlike existing models, which typically optimize only one or two objectives, HBES-BCO addresses all these factors simultaneously, providing a more comprehensive and efficient solution. Extensive simulations demonstrate that the HBES-BCO algorithm achieves a power loss is 5.4 W, and other models such as PTS-RA, IW-PSO, HLFO, and MoHHOTS given the power loss scores to be 10, 7.68, 5.25, 5.55, and 5.59, making it a highly effective and innovative solution for cloud-edge task scheduling. These improvements demonstrate the effectiveness of the hybrid optimization approach in

关键词：

来源：评论

学校读者我要写书评

暂无评论

Assessment of suitability of Harten’s total variation diminishing scheme for hypersonic flow problems

引用

AIP Conference Proceedings 2022年第1期2516卷

作者： Shouvik Bandopadhyay S. Senthilkumar Edison Chandraseelan 1Department of Aerospace Engineering SRM Institute of Science and Technology Kattankulathur Chennai Tamilnadu India 2Department of Mechanical Engineering AMET University Chennai Tamilnadu India

Since the advent of computational simulations for understanding physical phenomena, there has been a constant effort to develop better, more efficient and more accurate numerical schemes. In this regard, the Harten’s Total Variation Diminishing (TVD) Scheme is known to give solutions of higher accuracy even with coarse grids. In the present work, suitability of this scheme is assessed for a hypersonic flow problem. Explicit TVD scheme of first and second order accuracy with entropy fix is utilized to construct an in-house 2D solver to solve the Euler Equations of fluid flow in both the Eulerian and Lagrangian coordinate systems. Min-mod limiter function was utilized to avoid spurring oscillations while implanting the second-order method. The results are first validated for supersonic flow problem followed by assessments of suitability of the scheme for hypersonic flows with three inlet Mach numbers (M=5, 6 and 8). It is found that the scheme was highly effective for the supersonic flow case. However, for hypersonic flow problems, solutions couldn’t be obtained for inlet M=8 with the 2nd order scheme. This could be attributed by a fact that there are sudden Jacobian transformations due to local mesh density.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cerebrum Tumor Segmentation of High Resolution Magnetic Resonance Images Using 2D-Convolutional Network with Skull Stripping

引用

Neural Processing Letters 2020年第4期53卷 2567-2580页

作者： Pitchai, R. Madhu Babu, Ch Supraja, P. Challa, Mahesh Kumar Department of Computer Science and Engineering B.V. Raju Institute of Technology Narsapur Medak Dist India Department of Information Technology School of Computing SRM Institute of Science and Technology Chennai India

The automatic segmentation of the tumor region from Magnetic Resonance cerebrum imageries is a difficult task in medical image analysis. Numerous techniques have been created with the goal of improving the segmentation effectiveness of the automated framework. As of late, Convolutional Neural Networks have accomplished better performance in various recognition tasks. In this paper, 2D-ConvNet with skull stripping (SS-2D ConvNet) based brain tumor segmentation technique have been presented. In the proposed method, initially, the input MRI images are preprocessed to reduce noise and skull stripped to correct the contrast and non-uniformity. It is further processed through the 2D-ConvNet for the segmentation of brain tumor. In particular, the proposed method has been compared with other existing methods, and it achieves better performances and yield precise segmentation with dice scores of 91%, accuracy of 89%, specificity of 98%, and sensitivity of 87%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A study on convolutional neural networks with active video tubelets for object detection and classification 1

引用

International Conference on Soft Computing and Signal Processing, ICSCSP 2018

作者： Rajkumar, R. Arunnehru, J. Department of Computer Science and Engineering SRM Institute of Science and Technology Vadapalani Campus ChennaiTamil Nadu India

ISBN: (数字)9789811333934

ISBN: (纸本)9789811333927

Convolutional neural networks are a powerful learning model inspired from biological concept of neurons. This deep learning model allows us to replicate the complex neural structure seen in living beings to be applied on data sets and to structure convulsions consisting of several layers. A study on convolutional neural networks have been proven to be an effective class of models for object recognition, taking those results into consideration we intend to apply convolutional neural networks for video classification in two different ways. Generalization of the results obtained by the application of convolutional neural networks on existing data sets for videos, namely Sports 1-M and YouTube object data set (YTO) and their implementation of two distinct CNNs. © Springer Nature Singapore Pte Ltd. 2019.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Enhancing depression detection in social media using firefly and bee colony optimized support vector machines: A novel machine learning approach

引用

AIP Conference Proceedings 2024年第1期3193卷

作者： Ajay Singh Yadav G. Akilandasowmya G. Arvind Kumar Suryansh Ahlawat Ashok Kumar N. Meenatchi R. Premkumar V. Vijayan Department of Mathematics SRM Institute of Science and Technology Delhi-NCR Campus Ghaziabad Uttar Pradesh 201204 India Department of Computer and Communication Engineering Sri Sairam Institute of Technology Chennai Tamil Nadu 600044 India Department of Electronics and Telecommunication Engineering Ramaiah Institute of Technology Bangalore Karnataka 560054 India Department of Computer Science and Engineering (Data Science) SRM Institute of Science and Technology Delhi-NCR Campus Ghaziabad Uttar Pradesh 201204 India Department of Computer Science Banasthali Vidyapith Banasthali Rajasthan 304022 India Department of Mathematics R.M.K. Engineering College Kavarapettai Tamil Nadu 601206 India Department of Electronics and Instrumentation Engineering Sri Sairam Engineering College Chennai Tamil Nadu 600044 India Department of Mechanical Engineering K. Ramakrishnan College of Technology Trichy Tamilnadu India

One of the most important places for people to share their thoughts, feelings, and opinions in this age of digital communication is on social media platforms. Depression and other mental health issues have become more prominent in this wave of user-generated content because of the widespread effects they have on people’s lives and on society as a whole. Its negative effects can be better managed and prevented if caught and treated early. Depressive symptoms in social media feeds can be detected with the use of a new ML textual data classification framework that combines a Fire fly Optimal (FFO) using Support Vector Machine (SVM) with an Artificial Bee Colony (ABC) Optimal using SVM classifier. Data cleansing and collection come first, subsequent by feature extractions through Time Frequencies Inversed Documented Frequency (TFIDF). After features are extracted, classification is done with FFOSVM and ABCOSVM. The Support Vector Machine parameters can be fine-tuned with the help of Fire fly Optimization and Artificial Bee Colony. The proposed framework strengthens classification accuracy by utilizing ML and the optimization capabilities of Fire fly Optimization and Artificial Bee Colony Algorithm. To measure the effectiveness of the framework, we conduct thorough experiments and analyses. To manage the intricacy of identifying depression from social media data, the results showed that the proposed classification algorithms performed better than standard methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Installation of IR camera in smart watches for facial expression recognition

Installation of IR camera in smart watches for facial expres...

引用

International Conference on Physics and Photonics Processes in Nano sciences 2019

作者： Sathish, K. Mojamill, Shaikh Priya, G. Sai Krishna, Raguram B. S. Sahoo, Sapna Department of Computer Science and Engineering SRM Institute of Science and Technology Ramapuram Chennai Tamil Nadu600089 India

IR camera, with its various advantages, can be found in many different fields. The most important application of an IR camera is that it can be used even in the dark, unlike an ordinary camera. Using this plus point, we can integrate this with other technologies to help us locate obstacles in the dark. This is due to the fact that IR cameras can detect heat instead of just capturing and since everything on this planet gives off heat, these IR cameras can detect these obstacles. Many methods have been implemented to deal with such problems but there are not easily available and are quite expensive. IR camera have been used in as surveillance cameras and also in movies to make the inanimate idealistic. However, they have not been integrated with smart watches which are easily available nowadays and are much cheaper than any other alternatives. The idea is to integrate this IR camera with a smart watch that would make it portable and easy to use anywhere. Not only is this technology aimed to help the general public but also the blind. Detection of any facial expression would alert the watch wearer that someone is in vicinity. Since IR camera can sense heat, they would be better recognizers of different expressions. This would not only allow them to navigate on their own, but also inform about the whereabouts of the user to their close friends and family in case of emergency. © Published under licence by IOP Publishing Ltd.

关键词： Watches

来源：评论

学校读者我要写书评

暂无评论

Evaluating oil spill contamination around Qaruh Island, Kuwait: Insights from EOS-04 C-band SAR data

引用

Regional Studies in Marine science 2025年 89卷

作者： N. Krishnaraj Mohammad Alamgeer Shouki A. Ebad Haya Mesfer Alshahrani Saraswathi A. Sumaiya Begum J. Jayanthi Department of Networking and Communications School of Computing College of Engineering and Technology SRM Institute of Science and Technology Kattankulathur- 603203 Tamil Nadu India Department of Information Systems College of Science & Art at Mahayil King Khalid University Saudi Arabia Department of Computer Science Faculty of Science Northern Border University Arar Saudi Arabia Department of Information Systems College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh Saudi Arabia Department of Computer Science and Engineering Chennai Institute of Technology Chennai Tamil Nadu India Department of Electronics and Communication Engineering R.M.D.Engineering College Tamil Nadu India Department of Computer Science and Engineering Sona College of Technology Salem Tamil Nadu India

Oil spills threaten marine environments, especially in ecologically sensitive regions like Qaruh Island, Kuwait. This study investigates the extent and severity of oil contamination surrounding Qaruh Island by leveraging C-Band Synthetic Aperture Radar (SAR) data acquired from two key satellite sources, Sentinel-1A and EOS-04, on April 23, 2024. Both satellites provided dual-polarized SAR data (HH and VV), which were analysed to enhance oil spill detection through polarization-based discrimination techniques. The analysis compared backscatter values from oil-covered and clean water surfaces to delineate the spill area accurately. Sentinel-1A and EOS-04 SAR data were pre-processed for geometric and radiometric corrections, followed by the application of filtering techniques to mitigate speckle noise and enhance detection clarity. Through this approach, we successfully detected a continuous oil spill extending over an area of approximately 20 square kilometres. The HH polarization was particularly effective in identifying the boundaries of the spill, while VV polarization contributed to confirming the presence of oil by highlighting differences in surface roughness between contaminated and uncontaminated areas. The results demonstrate that dual-polarized SAR data, specifically C-Band data, are crucial in oil spill monitoring, providing timely and accurate contamination assessments in coastal and marine environments. By utilizing Sentinel-1A and EOS-04 SAR datasets, this study underscores the effectiveness of multi-source SAR data for rapid response in environmental monitoring, ultimately supporting resource protection and mitigation efforts around Qaruh Island.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Behaviour-constrained support vector machines for FMRI data analysis

引用

AIP Conference Proceedings 2023年第1期2523卷

作者： Pardeep Kumar J. Jegan Amarnath K. Somasundaram Thilagavathy A. 1Department of Artificial Intelligence Anurag University Hyderabad Telangana India. 2Department of Computer Science and Engineering Sri Sairam Engineering College Chennai Tamil Nadu India. 3Department of Computer Science and Engineering Chennai Institute of Technology Malayambakkam Chennai Tamil Nadu India. 4Department of Computer Science and Engineering R.M.K. Engineering College Chennai Tamil Nadu India.

Statistical learning approaches, including SVM, are a powerful tool for studying functional imaging data. These were especially effective in extracting information from useful magnetic resonance imaging signals from the direction and location of visual properties and categories of objects. Computational intelligence methods are developed to decipher neural imagery information as a valuable tool. The distorted signal and the small number of test patterns usually reported in functional brain imaging studies pose problems when interpreting brain data by using predictive learning techniques. To solve this issue, we propose to use previous details focused on human observers’ behavioral success to boost the training of Support Vector Machines (SVMs). During functional resonance screening, we gather behavioral reactions to human participants executing a classified mission. We use the brain imaging method developed based on the behavioral choices of observers as a distance limit for SVM training. We name the behaviourally limited SVM process. Our results show that BCSVM consistently outperforms SVM.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Predicting the error in student's grades using an outlier approach

引用

AIP Conference Proceedings 2023年第1期2492卷

作者： K. Batri M. Sivaram Ahmed Najat Ahmed P. Saranya PSNA College of Engineering and Technology Dindigul India. Department of Information Technology MLR Institute of Technology Hyderabad Telangana India. Department of Computer Engineering Lebanese French University KR Iraq. Department of Computer Science and Engineering SRM Institute of Science and Technology Chennai India – 603203.

In this article, we propose an error prediction model. Our model predicts error in the student's grades. The outlier approach becomes the basis for the proposed error prediction model. The outlier approach uses Mean Median, and other statistical parameters. Apart from that, the distance calculation is also used to predicts the outlier. To calculate the distance, the records are treated as a multi-dimensional vector. But these existing approaches cannot be used to predict the error in the students’ grades, as the range of marks constitutes the grades. We propose this prediction model based on one basic assumption. Student's performance across all subjects will remain similar, and there won't be a drastic variation in grades. The proposed model is tested over the grades of the university end-semester examination. The maximum accuracy level of our proposed model stood around 90%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

IoT enabled smart sensor network to monitor the accessibility of fire extinguishers and fire exits around the factory

引用

AIP Conference Proceedings 2024年第1期2853卷

作者： S. Prasanna Bharathi R. Abishek M. Vignesh Sundar V. S. Thinesh Department of Electronics Engineering Vellore Institute of Engineering Vellore Tamil Nadu India Department of Electronics and Communication Engineering College of Engineering & Technology SRM Institute of Science and Technology Vadapalani Campus No 1 Jawaharlal Nehru Road Vadapalani Chennai - 600026 Tamil Nadu India

According to statistics from the Labour and Employment Ministry of India, for the years 2014-2016, forty-seven workers were injured and three died in accidents every day. A total of 3,562 workers have lost their lives while 51,124 were injured in accidents that occurred in factories across the country during that period. Industrial safety is pivotal for any manufacturing industry to prevent or lessen workplace injury. Fire extinguishers and fire exits are the two most important safety measures in case of a fire in the factory. But in most situations, accessibility to fire extinguishers and fire exits is blocked by cardboards, broken chairs, and other industrial waste that employees put there unknowingly. The main reason for this behavior is that fire extinguishers and fire exits are placed in open areas with large spaces so that accessibility to them would be easier. This has enabled employees to utilize the large open space to dump waste. To avoid these problems, we have developed an IoT-based smart sensor network to monitor fire extinguishers and fire exits for any obstacle that is blocking them. If there are any violations, the safety team will be notified of the location by the system.

关键词：

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：