检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Li, Hongkang Zhang, Shuai Zhang, Yihua Wang, Meng Liu, Sijia Chen, Pin-Yu The Dept. of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute United States New Jersey Institute of Technology United States The Dept. of Computer Science and Engineering Michigan State University United States IBM Research United States

Group imbalance has been a known problem in empirical risk minimization (ERM), where the achieved high average accuracy is accompanied by low accuracy in a minority group. Despite algorithmic efforts to improve the minority group accuracy, a theoretical generalization analysis of ERM on individual groups remains elusive. By formulating the group imbalance problem with the Gaussian Mixture Model, this paper quantifies the impact of individual groups on the sample complexity, the convergence rate, and the average and group-level testing performance. Although our theoretical framework is centered on binary classification using a one-hidden-layer neural network, to the best of our knowledge, we provide the first theoretical analysis of the group-level generalization of ERM in addition to the commonly studied average generalization performance. Sample insights of our theoretical results include that when all group-level co-variance is in the medium regime and all mean are close to zero, the learning performance is most desirable in the sense of a small sample complexity, a fast training rate, and a high average and group-level testing accuracy. Moreover, we show that increasing the fraction of the minority group in the training data does not necessarily improve the generalization performance of the minority group. Our theoretical results are validated on both synthetic and empirical datasets, such as CelebA and CIFAR-10 in image classification. Copyright © 2024, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Multivariate Optimal Hybrid Deep Learning Model for Forecasting of Day-Ahead Solar Irradiance with Meteorological Constraints

Multivariate Optimal Hybrid Deep Learning Model for Forecast...

引用

Annual North-American Power Symposium

作者： Sohag Kumar Saha Satish M. Mahajan Dept. of Electrical & Computer Engineering (ECE) Tennessee Technological University Center for Energy Systems Research (CESR) Tennessee Technological University Cookeville TN USA

ISBN: (数字)9798331521035

ISBN: (纸本)9798331521042

With the growing integration of solar power gener-ation into smart grids, accurate solar irradiance forecasting is of paramount importance for efficient grid operation and renew-able energy management. In this paper, a data decomposition approach with two different methods of deep learning and grid search optimization was combined to develop a better hybrid model to forecast solar irradiance. The proposed model combines spatial and temporal information to improve forecasting accuracy by considering the impact of meteorological constraints such as global horizontal irradiance, temperature, relative humid-ity, wind speed, cloud type, direct normal irradiance, diffuse horizontal irradiance, and solar zenith angle. In addition to the model architecture, this work incorporates hyper-parameter optimization to fine-tune the parameters of the model for optimal performance. The design of the model ensures that the system adapts to the specific characteristics of the solar irradiance data and meteorological conditions under consideration. The proposed hybrid model was evaluated using real-world data to outperform traditional forecasting methods in terms of accuracy. The results of the proposed hybrid model indicate better prediction ability as measured by four parameters (lower RMSE and MAE, fewer epochs, and a higher R2 co-efficient).

关键词： Solar irradiance Deep learning Adaptation models Accuracy Predictive models Data models Smart grids Forecasting Optimization Load modeling

来源：评论

学校读者我要写书评

暂无评论

Denoising and Extraction of Partial Discharge Pulse Characteristics Using Wavelet Denoising and Local Maxima Method

Denoising and Extraction of Partial Discharge Pulse Characte...

引用

National Proceedings Power and Energy Conference (PECon)

作者： Maninder Choudhary Ivo Palu Ivar Kiitam Muhammad Shafiq Paul Taklaja Abhinav Bhattarai Dept. of Electrical Power Engineering & Mechatronics Tallinn University of Technology Tallinn Estonia Center for Advanced Power Systems Florida Stage University Florida USA Dept. of Electrical Power Engineering & Mechatronics Khatmandu University Khatmandu Nepal

ISBN: (数字)9798350361971

ISBN: (纸本)9798350361988

Partial discharge (PD) pulse characteristics describe the nature of an insulation defect, its severity, and aging behavior of electrical insulation. However, extracting these characteristics requires sophisticated tools in order to denoise measured PD data, detect actual PD pulses, and analyze the corresponding characteristics. This paper proposes an effective tool that utilizes a simple hybrid approach based on wavelet denoising and the local maxima technique that denoises PD data acquired using a high-frequency current transducer. After de noising, it detects PD pulses from a time series dataset and identifies corresponding primary features like peak pulse amplitude and location. In addition, it can extract pulse shape (rise time, fall time and pulse width), and pulse sequence (voltage interval and time interval) features. Developed tool is computationally efficient as it precisely detects maj or PD events and corresponding features from a large data set while rejecting noise. Efficiency of tool is checked by implementing it on various data sets obtained from experimental investigation performed on three distinct types of defects, i.e., corona, surface, and internal discharge. Results show that this tool can be used for the identification of defect types, their progression, and aging studies to estimate the remaining lifetime based on changing PD behavior.

关键词： Partial discharges Insulation Transducers Surface waves Surface discharges Noise reduction Time series analysis Aging Feature extraction Data mining

来源：评论

学校读者我要写书评

暂无评论

SLIPT in Joint Dimming Multi-LED OWC systems with Rate Splitting Multiple Access

SLIPT in Joint Dimming Multi-LED OWC Systems with Rate Split...

引用

IEEE International Conference on Communications (ICC)

作者： Sepideh Javadi Sajad Faramarzi Farshad Zeinali Hosein Zarini Mohammad Robat Mili Panagiotis D. Diamantoulakis Eduard Jorswieck George K. Karagiannidis Pasargad Institute for Advanced Innovative Solutions (PIAIS) Tehran Iran Dept. of Electrical Engineering Iran University of Science and Technology Tehran Iran Dept. of Computer Engineering Sharif University of Technology Tehran Iran Dept. of Electrical and Computer Engineering Aristotle University of Thessaloniki Thessaloniki Greece Institute for Communications Technology Technische Universitat Braunschweig Braunschweig Germany Artificial Intelligence & Cyber Systems Research Center Lebanese American University (LAU) Lebanon

ISBN: (数字)9781728190549

ISBN: (纸本)9781728190556

Optical wireless communication (OWC) systems with multiple light-emitting diodes (LEDs) have recently been explored to support energy-limited devices via simultaneous lightwave information and power transfer (SLIPT). The energy consumption, however, becomes considerable by increasing the number of incorporated LEDs. This paper proposes a joint dimming (JD) scheme that lowers the consumed power of a SLIPT-enabled OWC system by controlling the number of active LEDs. We further enhance the data rate of this system by utilizing rate splitting multiple access (RSMA). More specifically, we formulate a data rate maximization problem to optimize the beamforming design, LED selection and RSMA rate adaptation that guarantees the power budget of the OWC transmitter, as well as the quality-of-service (QoS) and an energy harvesting level for users. We propose a dynamic resource allocation solution based on proximal policy optimization (PPO) reinforcement learning. In simulations, the optimal dimming level is determined to initiate a trade-off between the data rate and power consumption. It is also verified that RSMA significantly improves the data rate.

关键词： Wireless communication Energy consumption Quality of service Reinforcement learning Light emitting diodes Dynamic scheduling Real-time systems

来源：评论

学校读者我要写书评

暂无评论

RadioModRec-l: A Large Scale Radio Frequency Dataset for Automatic Modulation Recognition Research

RadioModRec-l: A Large Scale Radio Frequency Dataset for Aut...

引用

International Conference on Emerging Trends in Networks and computer Communications (ETNCC)

作者： Emmanuel Adetiba Jamiu R. Olasina Obiseye Obiyemi Dept. of Electrical and Information Engineering Covenant Applied Informatics and Communication Africa Center of Excellence Covenant University Ota Nigeria HRA Institute for Systems Science Durban University of Technology Durban South Africa Dept. of Computer Engineering Federal Polytechnic Ilaro Ilaro Nigeria Dept. of Electrical Power Engineering Space Science Centre Durban University of Technology Durban South Africa

ISBN: (数字)9798350353266

ISBN: (纸本)9798350353273

Access to well-curated large datasets remains a significant bottleneck in AI-based research within wireless communication. Rapid advancements in neighbouring fields, such as computer vision and natural language processing, are largely due to the availability of extensive open-access datasets. However, similar progress has not been observed in wireless communication. To address this gap, we curated a comprehensive dataset for fifteen digital modulation schemes, including 4QAM, 16QAM, 64QAM, 256QAM, 8PSK, 16PSK, 32PSK, 64PSK, 128PSK, 256PSK, CPFSK, DBPSK, DQPSK, GFSK, and GMSK. Our dataset considers Rayleigh and Rician channel models under Additive White Gaussian Noise (A WGN) with SNRs ranging from -20dB to + 20dB in 5dB increments. The data samples were converted to constellation signal images and carefully pre-processed. Named RadioModRec, this dataset provides a valuable resource for researchers to train and evaluate AI models. It is freely accessible on Kaggle, promoting further innovation in the wireless communication domain.

关键词： Wireless communication Technological innovation Transfer learning Stochastic processes Rician channels Market research Natural language processing Optimization Residual neural networks Radio spectrum management

来源：评论

学校读者我要写书评

暂无评论

Assessing Respiratory Changes Induced by STN-DBS in a Rat Model 19

Assessing Respiratory Changes Induced by STN-DBS in a Rat Mo...

引用

19th International Conference on Emerging Technologies, ICET 2024

作者： Inam, Shehla Teixeira, Livia De Mesquita Waqar, Muhammad Ishaq, Muhammad Iqbal, Shahzaib Ullah, Fasih Muthuswamy, Jitendran School of Biological & Health Systems Engineering Arizona State University Tempe United States Balochistan University of Engineering & Technology Dept. of Biomedical Engineering Khuzdar Pakistan National University of Computer & Emerging Sciences Dept. of Electrical Engineering Islamabad Pakistan Balochistan University of Engineering & Technology Dept. of Mechanical Engineering Khuzdar Pakistan Cecos University of It & Emerging Sciences Dept. of Electrical Engineering Peshawar Pakistan

ISBN: (纸本)9798350379587

Deep brain stimulation of the subthalamic nucleus (STN-DBS) is a technique used to treat motor symptoms of movement disorders including Parkinson's Disease (PD), a neurogenerative disorder caused by the progressive loss of striatal dopamine in the brain that leads to tremors, rigidity, and postural instability. Despite its therapeutic effects, DBS is still known to cause some non-motor symptoms including breathlessness due to unknown reasons in PD patients. In this paper, the changes in breathing rates were evaluated from peak detection for 20 trials in the time domain for the preliminary data obtained from a rat model after the application of DBS on the STN for frequencies 10Hz and 130Hz with stimulation intensities of 150 μA and 450 μA. Results revealed a significant change in breathing rates for all intensities and corresponding frequencies, confirming that the degree of disruption in breathing rate may depend on the stimulation frequency. The study may further suggest potential neuronal networks that mediate these changes. © 2024 IEEE.

关键词： Neurodegenerative diseases

来源：评论

学校读者我要写书评

暂无评论

LEARNING EFFICIENT POSITIONAL ENCODINGS WITH GRAPH NEURAL NETWORKS

arXiv

引用

arXiv 2025年

作者： Kanatsoulis, Charilaos I. Choi, Evelyn Jegelka, Stephanie Leskovec, Jure Ribeiro, Alejandro Dept. of Computer Science Stanford University United States School of Computation Information and Technology Technical University of Munich Computer Science & Artificial Intelligence Laboratory Massachusetts Institute of Technology United States Dept. of Electrical and Systems Engineering University of Pennsylvania United States

Positional encodings (PEs) are essential for effective graph representation learning because they provide position awareness in inherently position-agnostic transformer architectures and increase the expressive capacity of Graph Neural Networks (GNNs). However, designing powerful and efficient PEs for graphs poses significant challenges due to the absence of canonical node ordering and the scale of the graph. In this work, we identify four key properties that graph PEs should satisfy: stability, expressive power, scalability, and genericness. We find that existing eigenvector-based PE methods often fall short of jointly satisfying these criteria. To address this gap, we introduce PEARL, a novel framework of learnable PEs for graphs. Our primary insight is that message-passing GNNs function as nonlinear mappings of eigenvectors, enabling the design of GNN architectures for generating powerful and efficient PEs. A crucial challenge lies in initializing node attributes in a manner that is both expressive and permutation equivariant. We tackle this by initializing GNNs with random node inputs or standard basis vectors, thereby unlocking the expressive power of message-passing operations, while employing statistical pooling functions to maintain permutation equivariance. Our analysis demonstrates that PEARL approximates equivariant functions of eigenvectors with linear complexity, while rigorously establishing its stability and high expressive power. Experimental evaluations show that PEARL outperforms lightweight versions of eigenvector-based PEs and achieves comparable performance to full eigenvector-based PEs, but with one or two orders of magnitude lower complexity. Our code is available at https://***/ehejin/Pearl-PE. Copyright © 2025, The Authors. All rights reserved.

关键词： Encoding (symbols)

来源：评论

学校读者我要写书评

暂无评论

The Impact of the Energy Storage System Operation Strategy on Power Grid Reliability

引用

Transactions of the Korean Institute of electrical Engineers 2021年第12期70卷 1820-1826页

作者： Kim, Beopsoo Lee, Dayoung Rusetskii, Nikita Shusterzon, Konstantin Sidorov, Denis Kim, Insu Dept. of Electrical Engineering Inha University Korea Republic of Dept. of Electrical and Computer Engineering Inha University Korea Republic of School of Information Technology and Data Science Irkutsk National Research Technical University Russia Energy Systems Institute Siberian Branch Russian Academy of Sciences Russia

Based on the empirical data published by the Korea Electric Power Exchange and the Korea Meteorological Administration, we present a methodology that predicts the system marginal prices. For this purpose, this study uses the long-short term memory (LSTM) technique, which is commonly used in time-series prediction problems. As a result, we present a strategy for operating energy storage systems (ESSs) from the perspective of a prosumer rather than a system operator or conventional utilities. This study also examines the effects of the proposed prosumer actions on the system. The proposed operating strategy is established based on dynamic programming. In particular, this study takes the proposed operation strategy of ESSs the point of view of system reliability. © The Korean Institute of electrical Engineers This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://***/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

关键词： Reliability

来源：评论

学校读者我要写书评

暂无评论

EfficientNetB0-based Rice Disease Classification: A Strategic Fine-Tuning Approach

EfficientNetB0-based Rice Disease Classification: A Strategi...

引用

Smart Electronics and Communication (ICOSEC), International Conference on

作者： Shivani Mishra Chavali Sai Sreeram Yadav Akansha Raju A Balasundaram Ayesha Shaik Dept. of Computer Science and Engineering Vellore Institute of Technology Chennai Chennai India Center for Cyber Physical Systems Vellore Institute of Technology Chennai Chennai India

ISBN: (数字)9798331504403

ISBN: (纸本)9798331504410

Rice, a major stape food, is crucial for both sustenance and the living of millions of farmers. Detecting rice leaf diseases is vital to prevent widespread crop losses, safeguarding farmers’ incomes and global food supply. While current detection systems face accuracy and robustness challenges, deep learning models, particularly the EfficientNetB0, show promise. Despite their sensitivity to noise and variations in leaf appearance, ongoing research addresses these issues. This study introduces the use of EfficientNetB0 for accurate classification, incorporating a fine-tuning strategy in a brief five-epoch training period. This approach optimally balances model adaptation and prevents overfitting, leveraging the deep neural network’s ability to capture complex hierarchical features. The limited fine-tuning duration addresses potential risks while maximizing the model’s robust features. This unique amalgamation of advanced architecture and tailored training strategy aims to achieve efficient and accurate disease classification. The research signifies a step toward revolutionizing rice disease detection, emphasizing the importance of striking a balance between cutting-edge methodologies and practical considerations in the pursuit of agricultural sustainability. Before Fine Tuning the Training Accuracy was $\mathbf{9 8. 4 6 \%}$, Validation Accuracy was $\mathbf{9 9. 1 5 \%}$ and Test Accuracy was $\mathbf{9 8. 4 4 \%}$.After Fine Tuning there was a significant increase in the accuracy. The Training Accuracy came to be $\mathbf{9 9. 7 7 \%}$, Validation Accuracy is $\mathbf{9 9. 4 3 \%}$ and Test Accuracy: 99.12%

关键词： Training Adaptation models Accuracy Sensitivity Computational modeling Noise Robustness Sustainable development Tuning Diseases

来源：评论

学校读者我要写书评

暂无评论

Joint Energy and Computation Workload Management for Geo-distributed Data centers

IEEE Transactions on Green Communications and Networking

引用

IEEE Transactions on Green Communications and Networking 2025年

作者： Wang, Ran Wu, Rixin Liu, Linfeng Yi, Changyan Zhu, Kun Wang, Ping Niyato, Dusit Nanjing University of Aeronautics and Astronautics Collaborative Innovation Center of Novel Software Technology and Industrialization College of Computer Science and Technology Nanjing China Nanjing University of Posts and Telecommunications School of Computer Science and Technology Nanjing China York University Dept. of Electrical Engineering and Computer Science Lassonde School of Engineering Canada Nanyang Technological University School of Computer Science and Engineering Singapore

The increasing demands of data computation and storage for cloud-based services motivate the development and deployment of large-scale data centers (DCs). The energy demand of these devices is rising rapidly and becoming a noticeable challenge for current power networks. The smart grid (SG) is deemed as the future power system paradigm enabling more affordable and sustainable energy supply, which can effectively relieve the load pressure from DCs. Moreover, with growing concerns regarding harmful emissions due to combustion of fossil fuels, the exploitation of renewable energy sources (RES) has attracted extensive attention, which can benefit SGs and DCs, as well as society at large. However, the geo-distributed property of DCs and SGs and the uncertain nature of RES production pose severe challenges to the optimal management of computation and energy resources in such a tripartite coupling system. Focusing on these issues, a joint energy and computation workload management framework is proposed for enabling a sustainable DC paradigm with distributed RES. Specifically, a three-layer game is formulated to model the iterations among entities including the energy market, data center operators (DCOs), and SGs. The market includes a certain amount of RES that must be dispatched. The SG offers the DCO an electricity selling price while simultaneously importing RES from the market at a buying price in order to maximize the benefit. The DCO allocates the workload to different DCs, aiming to minimize the costs of energy consumption and carbon emissions. The interactive processes between different entities are further decomposed into two coupling Stackelberg games. We obtain the equilibrium state of the game and prove its uniqueness and optimality. Simulation experiments are conducted to evaluate the performance of the joint energy and computation workload management scheme and show its superiority over counterparts in utilizing renewable energy and reducing emissions. Furthe

关键词： Smart power grids

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：