检索结果-内蒙古大学图书馆

Big Data Analytics and Practices (IBDAP), International Conference on

作者： Sakorn Mekruksavanich Ponnipa Jantawong Anuchit Jitpattanakul Department of Computer Engineering School of Information and Communication Technology University of Phayao Phayao Thailand Department of Mathematics Faculty of Applied Science Intelligent and Nonlinear Dynamic Innovations Research Center King Mongkut’s University of Technology North Bangkok Bangkok Thailand

Epilepsy, a neurological disorder, is typically identified by analyzing electroencephalogram (EEG) signals. However, manually inspecting epileptic seizures is a time-consuming and labor-intensive task. Previous automatic detection algorithms utilizing traditional methods have demonstrated good accuracy in specific EEG classification scenarios but have shown poor performance in others. To tackle this challenge, our study introduces a new deep learning model called the self-normalizing neural network. This model leverages convolutional neural networks and recurrent neural networks to enhance the detection of epileptic seizures using EEG signals. To evaluate our approach, we designed and assessed the proposed detection model using a publicly available Epileptic Seizures Recognition Dataset (ESRD). The experimental findings indicate that the self-normalizing neural network achieves a performance improvement of up to 1.03% accuracy and 1.78% F1-score in epileptic seizure detection, surpassing traditional convolutional neural network approaches.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Attention Based Hybrid Deep Learning Network for Locomotive Mode Recognition in Natural Environments Using Wearable Sensors

Attention Based Hybrid Deep Learning Network for Locomotive ...

引用

Research, Invention, and Innovation Congress (RI2C)

作者： Sakorn Mekruksavanich Ponnipa Jantawong Anuchit Jitpattanakul Department of Computer Engineering School of Information and Communication Technology University of Phayao Phayao Thailand Department of Mathematics Intelligent and Nonlinear Dynamic Innovations Research Center Faculty of Applied Science King Mongkut's University of Technology North Bangkok Bangkok Thailand

The field of pervasive computing encompasses the significant domain of study known as WS-LMR, which focuses on identifying locomotive modes through wearable sensors. WS-LMR aims to automatically evaluate and interpret individual locomotive actions along with contextual information using wearable sensor data. This technology finds applications in medical surveillance systems and various systems involving interactions between individuals and intelligent wearable devices. However, existing WS-LMR algorithms are often evaluated using data collected under controlled conditions, limiting their effectiveness in uncontrolled real-world environments. This scholarly article addresses the primary objective of identifying human locomotion activities in real-world contexts. To enhance the efficacy of the WS-LMR system, our study proposes the utilization of an attention-based hybrid deep learning network that combines convolutional neural network and long-short term memory networks, referred to as the Att-CNN-LSTM model. We evaluated the performance of the proposed network using the publicly available Bath Natural Environment HAR dataset, a benchmark LMR dataset. Both model training and testing were conducted using a 5-fold cross-validation approach. The Att-CNN-LSTM model achieved the highest accuracy of 94.25 % and an F1-score of 91.47% based on the results obtained from multiple experiments.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Incentivizing or Disincentivizing Participation in Demand Response Programs? Estimation of Customer Willingness to Accept

SSRN

引用

SSRN 2025年

作者： Wang, Yu Howard, Gregory Zhang, Wendong Brown, Marilyn Morton, Abigail Cetin, Kristen McCalley, James Department of Political Science Iowa State University United States Department of Economics East Carolina University United States The Charles H. Dyson School of Applied Economics and Management Cornell University United States School of Public Policy Georgia Institute of Technology United States Federal Energy Regulatory Commission United States Department of Civil and Environmental Engineering Michigan State University United States Department of Electrical and Computer Engineering Iowa State University United States

Demand response has great potential to provide value flexibility services to the grid, which can be harnessed through direct load control (DLC) programs. Many utilities have provided direct financial incentives to encourage residential customers to enroll in DLC programs. However, financial incentives do not always result in high participation rate. This study aims to assess residential customers’ willingness to enroll in utility-run DLC programs and analyze how incentives affect participation intention. A survey using the contingent valuation method (CVM) collected a random sample of 591 residents in the U.S. Midwest from 2020-2022, which finds, overall, 56.7% of residents are willing to enroll in DLC. It is estimated that 61.5% of residents are willing to participate due to intrinsic motivations (i.e., without monetary incentive), while a small token ($15-24 sign-on bonus) only leads to about 50% of the customers agreeing to enroll. To reach the same level of participation from intrinsic motivations, utilities must offer a much higher incentive - $83-$90. Additionally, we find several factors affect acceptance of DLC programs, including knowledge learning, concerns of comfort loss, and personal beliefs on energy savings, reliability, and environmental benefit. Findings from this study have significant policy implications and help improve the design of utility DLC programs. When developing strategies to recruit participants, no mention of monetary incentives could increase the participation of resident customers. We also recommend that utilities consider educational and demonstration programs to engage stakeholders, and information campaigns to emphasize the potential benefits without compromising thermal comfort. © 2025, The Authors. All rights reserved.

关键词： Sales

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Approaches for Epileptic Seizures Recognition based on EEG Signal

Deep Learning Approaches for Epileptic Seizures Recognition ...

引用

International Conference on Telecommunications and Signal Processing (TSP)

作者： Sakorn Mekruksavanich Anuchit Jitpattanakul Department of Computer Engineering School of Information and Communication Technology University of Phayao Phayao Thailand Department of Mathematics Faculty of Applied Science Intelligent and Nonlinear Dynamic Innovations Research Center King Mongkut’s University of Technology North Bangkok Bangkok Thailand

Recognition system of human actions is a popular research field in healthcare which is essential in identifying abnormal patient activities to estimate their psychological state. Epileptic seizures are a common neurological disorder affecting millions of individuals globally. Early and accurate diagnosis of epilepsy can lead to up to 70% of seizure-free patients. There is a requirement for smart and automated HAR systems to aid clinicians in accurately diagnosing neurological disorders. This study examines the latest deep learning networks for time-series classification to detect epileptic seizures using EEG signal data. To enhance recognition performance, a hybrid residual network called Inception-ResNet is presented for the binary classification of epileptic activities. Through thorough analysis, the Inception-ResNet model was observed to provide the best results, with an average accuracy and F1-score of 99.61%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Position-aware Human Activity Recognition with Smartphone Sensors based on Deep Learning Approaches

Position-aware Human Activity Recognition with Smartphone Se...

引用

International Conference on Telecommunications and Signal Processing (TSP)

Numerous research studies on human activity recognition (HAR) have utilized smartphone sensors, such as accelerometers, gyroscopes, and magnetometers. The accelerometer, in particular, has been considerably explored as the primary sensor in HAR research. Recently, researchers have integrated different smartphone sensors to enhance recognition interpretation. Nevertheless, it is still challenging to determine the optimal usage of these sensors, either individually or in combination, for better identification effectiveness in various motion situations. In this article, we explore how different motion sensors behave during the activity recognition process, using various deep learning (DL) approaches to classify physical actions based on smartphone sensor data. We evaluate the DL models using a publicly benchmarked dataset of data collected from ten participants performing eight movements while carrying smartphones in different situations. Our experimental results reveal that, except for the magnetometer, each sensor can play a leading role in HAR, depending on the recognized activity, body position, data features employed, and classification approach. Additionally, we find that the combination of sensors only improves overall recognition interpretation when their individual performances are low.

关键词：

来源：评论

学校读者我要写书评

暂无评论

STAR-RIS-NOMA Assisted Power Constrained Covert Communication Systems

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Lei, Yuxin Huang, Gaojian Li, Xingwang Tian, Zhifa Mumtaz, Shahid Asif, Muhammad Henan Polytechnic University School of Physics and Electronic Information Engineering Jiaozuo454003 China City University of Macau Faculty of Data Science 999078 China Henan Polytechnic University School of Surveying and Land Information Engineering Jiaozuo454003 China Henan Polytechnic University School of Physics and Electronic Information Engineering Jiaozuo454099 China Southeast University National Mobile Communications Research Laboratory Nanjing210096 China Silesian University of Technology Department of Applied Informatics Gliwice Poland Department of Engineering Nottingham Trent University United Kingdom Jiangsu University School of Computer Science and Communication Engineering Zhenjiang China

In this paper, a covert communication system assisted by a simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) in non-orthogonal multiple access (NOMA) networks is proposed. When the maximum channel power gain from the transmitter in the main network to the users in adjacent networks is smaller than the threshold, the transmitter sends the NOMA signals to the users with the assistance of STAR-RIS. Moreover, channel estimation errors and imperfect successive interference cancellation (SIC) are considered in the system analysis. To characterize the covert performance, the detection error probability (DEP) and the average minimum detection error probability (AMDEP) are first derived. Then, the closed-form solutions of the outage probability (OP) for the users are obtained. Finally, the maximum effective covert rate (ECR) subject to the covertness constraint and non-interruption constraint is studied. The numerical simulation results show that: 1) the detection performance of Willie (warden) deteriorates as the channel estimation errors increase;2) when the transmitter power exceeds the threshold, communication in the main network is interrupted;3) under a certain transmit power, the maximum ECR increases with the increasing of the number of STAR-RIS reflection elements. © 2025 IEEE.

关键词： Channel estimation

来源：评论

学校读者我要写书评

暂无评论

Optical Teleportation using Electro-Optic Plasmonic Graphene Waveguide

Optical Teleportation using Electro-Optic Plasmonic Graphene...

引用

Conference on Lasers and Electro-Optics (CLEO)

作者： Montasir Qasymeh Muhammad Asjad Hichem Eleuch Electrical and Computer Engineering Department Abu Dhabi University UAE Department of Applied Physics and Astronomy University of Sharjah Sharjah UAE Institute for Quantum Science and Engineering Texas AM University Texas USA

We demonstrate that optical teleportation can be realized by using two interacting optical fields in an electrically driven graphene waveguide. The simulations show that the proposed system can achieve high-fidelity t... 详细信息

ISBN: (数字)9781957171050

ISBN: (纸本)9781665466660

关键词： Optical losses Graphene Waveguide lasers Teleportation Plasmons Optical pumping Optical coupling

来源：评论

学校读者我要写书评

暂无评论

Predictive Auto-scaler for Kubernetes Cloud

Predictive Auto-scaler for Kubernetes Cloud

引用

Annual IEEE Systems Conference

作者： Simon Shim Ankit Dhokariya Devangi Doshi Sarvesh Upadhye Varun Patwari Ji-Yong Park Applied Data Science Department San Jose State University San Jose California USA Computer Engineering Department San Jose State University San Jose California USA Korea Internet and Security Agency San Jose California USA

One of the basic requirements with adapting to cloud technology is to find an optimal resource allocation based on the dynamic workload. The default functioning of Kubernetes Horizontal Pod Auto-scaling in cloud is scaling of its pods only when the threshold of the cluster/application is crossed in order to adapt to increasing workload. Rather we want to deploy a proactive provisioning framework based on machine learning based predictions. We have demonstrated a novel deep learning framework based on a transformer in the area of dynamic workload predictions and showed how to apply the results to a custom auto-scaler in cloud. Our Framework builds time-series predictive models in machine learning such as ARIMA, LSTM, Bi-LSTM and transformer models. The dynamic scaling framework applies machine learning algorithms and presents recommendations to make proactive and smart decisions. Though the transformer model has been used in NLP and Vision applications mostly, we showed that the transformer based model can produce the most effective results in cloud workload predictions.

关键词： Deep learning Measurement Machine learning algorithms Computational modeling Heuristic algorithms Supervised learning Predictive models

来源：评论

学校读者我要写书评

暂无评论

Improving Lower Limb Activity Recognition Based on Inertial Sensors Using CNN-LSTM Network

Improving Lower Limb Activity Recognition Based on Inertial ...

引用

Research, Invention, and Innovation Congress (RI2C)

作者： Sakorn Mekruksavanich Ponnipa Jantawong Anuchit Jitpattanakul Department of Computer Engineering School of Information and Communication Technology University of Phayao Phayao Thailand Department of Mathematics Intelligent and Nonlinear Dynamic Innovations Research Center Faculty of Applied Science King Mongkut's University of Technology North Bangkok Bangkok Thailand

The utilization of wearable sensors has gained significant popularity across various domains. This widespread adoption has driven the need for advancements in sensor technology, focusing on attributes such as lightweight design, dependability, and integration capabilities. One specific area where wearable sensor networks have faced challenges is in human activity recognition, particularly in accurately inferring movement data. The effectiveness of this process depends on several factors, such as the quality of sensor-derived features, the discriminatory power of selected features, and the chosen classification methodology. To address these challenges, this research introduces a novel hybrid deep learning architecture called CNN-LSTM. The primary goal of this architecture is to enhance the accuracy of recognizing human lower limb movement. To evaluate the effectiveness of the proposed model, a series of investigations were conducted using the publicly available HuGaDB dataset. This dataset consists of data collected from a body sensor network comprising six wearable inertial sensors, including accelerometers and gyroscopes. These sensors were positioned on both the right and left thighs, shins, and feet. The outcomes of the investigation demonstrated that the suggested model achieved an impressive precision of 97.81 % and a maximum Fl-score of 95.50%. These results highlight the effectiveness of the CNN-LSTM model in accurately recognizing human lower limb movement.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Edge AI: A Taxonomy, Systematic Review and Future Directions

arXiv

引用

arXiv 2024年

作者： Singh Gill, Sukhpal Golec, Muhammed Hu, Jianmin Xu, Minxian Du, Junhui Wu, Huaming Walia, Guneet Kaur Murugesan, Subramaniam Subramanian Ali, Babar Kumar, Mohit Ye, Kejiang Verma, Prabal Kumar, Surendra Cuadrado, Felix Uhlig, Steve School of Electronic Engineering and Computer Science Queen Mary University of London London United Kingdom Abdullah Gul University Kayseri Turkey Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China Center for Applied Mathematics Tianjin University Tianjin China Department of Information Technology National Institute of Technology Jalandhar India Department of Information Technology National Institute of Technology Srinagar India Department of Computer Engineering and Applications GLA University Mathura India Spain

Edge Artificial Intelligence (AI) incorporates a network of interconnected systems and devices that receive, cache, process, and analyze data in close communication with the location where the data is captured with AI technology. Recent advancements in AI efficiency, the widespread use of Internet of Things (IoT) devices, and the emergence of edge computing have unlocked the enormous scope of Edge AI. Edge AI aims to optimize data processing efficiency and velocity while ensuring data confidentiality and integrity. Despite being a relatively new field of research from 2014 to the present, it has shown significant and rapid development over the last five years. This article presents a systematic literature review for Edge AI to discuss the existing research, recent advancements, and future research directions. We created a collaborative edge AI learning system for cloud and edge computing analysis, including an in-depth study of the architectures that facilitate this mechanism. The taxonomy for Edge AI facilitates the classification and configuration of Edge AI systems while examining its potential influence across many fields through compassing infrastructure, cloud computing, fog computing, services, use cases, ML and deep learning, and resource management. This study highlights the significance of Edge AI in processing real-time data at the edge of the network. Additionally, it emphasizes the research challenges encountered by Edge AI systems, including constraints on resources, vulnerabilities to security threats, and problems with scalability. Finally, this study highlights the potential future research directions that aim to address the current limitations of Edge AI by providing innovative solutions. © 2024, CC BY.

关键词： Cloud platforms

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：