检索结果-内蒙古大学图书馆

arXiv 2023年

作者： LaBonte, Tyler Muthukumar, Vidya Kumar, Abhishek H. Milton Stewart School of Industrial and Systems Engineering Georgia Institute of Technology United States School of Electrical and Computer Engineering Georgia Institute of Technology United States Google DeepMind United Kingdom

Empirical risk minimization (ERM) of neural networks is prone to over-reliance on spurious correlations and poor generalization on minority groups. The recent deep feature reweighting (DFR) technique [33] achieves state-of-the-art group robustness via simple last-layer retraining, but it requires held-out group and class annotations to construct a group-balanced reweighting dataset. In this work, we examine this impractical requirement and find that last-layer retraining can be surprisingly effective with no group annotations (other than for model selection) and only a handful of class annotations. We first show that last-layer retraining can greatly improve worst-group accuracy even when the reweighting dataset has only a small proportion of worst-group data. This implies a "free lunch" where holding out a subset of training data to retrain the last layer can substantially outperform ERM on the entire dataset with no additional data or annotations. To further improve group robustness, we introduce a lightweight method called selective last-layer finetuning (SELF), which constructs the reweighting dataset using misclassifications or disagreements. Our empirical and theoretical results present the first evidence that model disagreement upsamples worst-group data, enabling SELF to nearly match DFR on four well-established benchmarks across vision and language tasks with no group annotations and less than 3% of the held-out class annotations. Our code is available at https://***/tmlabonte/last-layer-retraining. Copyright © 2023, 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

Addressing Non-stationarity with Stochastic Trend in the Context of Limited Time Series Data: An Experimental Survey

Research Square

引用

Research Square 2025年

作者： Batoure Bamana, Apollinaire Sokdou Bila Lamou, Yannick Fotsa-Mbogne, David Jaures Shafiee Kamalabad, Mahdi Department of Computer Engineering University Institute of Technology The University of Ngaoundere Cameroon Department of Robotic and Industrial Computing School of Chemical Engineering and Mineral Industries The University of Ngaoundere Cameroon Department of Mathematics and Computer Science Faculty of Science The University of Ngaoundere Cameroon Laboratory of Mathematics Computer Science and Applications Department of Mathematics and Computer Science National Higher School of Agro-Industrial Sciences The University of Ngaoundéré Cameroon Laboratory of Mathematics and Computer Science Department of Mathematics Statistics and Computer Science Faculty of Science The University of Bertoua Cameroon Department of Methodology and Statistics Faculty of Social and Behavioural Sciences Utrecht University Utrecht Netherlands Centre for Complex Systems Studies Utrecht University Utrecht Netherlands

Stationarity in time series is a key property for practical data analysis, inferences, and predictions particularly in biosciences. Stationarity can be either deterministic or stochastic. If a time series data is not stationary, it can be rendered stationary through detrending or differentiation techniques. Differentiation can be performed in either an integer or fractional form. This paper investigates non-stationarity in time series data, focusing specifically on stochastic trends characterised by a unit root. It hypothesises that fractional differentiation can enhance forecast accuracy for datasets with limited volume. For experiments, 24 series corresponding to weekly Malaria and Typhoid Fever from Adamawa Region (Cameroon) are analysed and forecasted. Each series comprises 156 observations covering January 2021 to December 2023. After collecting, checking and classifying the data, four forecasting models, Auto Regressive Integrated Moving Average (ARIMA), Fractional ARIMA (ARFIMA), Long Short-Term Memory (LSTM), and the proposed Fractional Differencing and LSTM (FD-LSTM), are implemented. The models' performances are evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of determination (R2) metrics, which inform further analysis and recommendations. Results reveal that ARFIMA outperformed the ARIMA by 93% in training and 100% in testing. Similarly, FD-LSTM achieved a 100% improvement over LSTM in training and testing. These findings underscore the value of achieving stationarity through fractional differentiation, enhancing model performance and providing a robust framework for forecasting in datasets with limited observations. © 2025, CC BY.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Global scale solar energy harnessing: An advanced intra-hourly diffuse solar irradiance predicting framework for solar energy projects

引用

Neural Computing and Applications 2024年第18期36卷 10585-10598页

作者： El-kenawy, El-Sayed M. Bailek, Nadjem Bouchouicha, Kada Zerouali, Bilel Hassan, Muhammed A. Kuriqi, Alban Jamil, Basharat Colak, Ilhami Khalil, Adel Ibrahim, Abdelhameed Department of Communications and Electronics Delta Higher Institute of Engineering and Technology Mansoura35111 Egypt Laboratory of Mathematics Modeling and Applications Department of Mathematics and Computer Science Faculty of Sciences and Technology Ahmed Draia University of Adrar Adrar01000 Algeria Energies and Materials Research Laboratory Faculty of Sciences and Technology University of Tamanghasset Tamanrasset10034 Algeria MEU Research Unit Middle East University Amman Jordan Adrar01000 Algeria Vegetal Chemistry-Water-Energy Laboratory Faculty of Civil Engineering and Architecture Department of Hydraulic Hassiba Benbouali University of Chlef Ouled Fares Chlef02180 Algeria Mechanical Power Engineering Department Faculty of Engineering Cairo University Giza12613 Egypt School of Architecture University of Waterloo 7 Melville St S CambridgeONN1S 2H4 Canada CERIS Instituto Superior Tecnico Universidade de Lisboa Lisbon Portugal Civil Engineering Department University for Business and Technology PristinaKosovo10000 High Temperature Processes Unit IMDEA Energy Institute Av. Ramón de La Sagra 3 Móstoles Madrid28935 Spain Faculty of Engineering and Architectures Nisantasi University Istanbul Turkey Computer Engineering and Control Systems Department Faculty of Engineering Mansoura University Mansoura35516 Egypt

Diffuse horizontal irradiance (DHI) forecasts are critical for adopting solar photovoltaic technology. Yet, they can lack reliability given the limited and uncertain meteorological data available for desert areas. This research develops a new sine cosine dipper throated optimization (SCDTO) technique, targeting DHI prediction even given such data constraints. SCDTO uniquely hybridizes sine cosine metaheuristics, adept at exploration, with dipper throat optimization, providing focused exploitation. This hybridization aims to create a robust ensemble model capable of delivering reliable DHI predictions despite climatic uncertainty. The ensemble model, employing various input combinations, was rigorously evaluated across multiple meteorological stations. The SCDTO algorithm exhibited remarkable performance improvements, yielding substantial reductions exceeding 93% in root mean squared error and 98% in mean absolute error at key stations in Morocco, including Tan-Tan, Zagora, Erfoud, and Oujda. Comparative analyses against established optimization algorithms consistently underscored the superior predictive capabilities of SCDTO. Visualizations, including box plots and histograms, demonstrated SCDTO’s efficacy in minimizing prediction errors, particularly for Tan-Tan, Zagora, and Erfoud stations. Also, statistical validation through one-way analysis of variance (ANOVA) further affirmed the significance of the proposed SCDTO method. Therefore, this hybrid metaheuristic optimization enables more accurate DHI predictions from limited meteorological data. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Solar energy

来源：评论

学校读者我要写书评

暂无评论

ECADA: An Edge Computing Assisted Delay-Aware Anomaly Detection Scheme for ICS

ECADA: An Edge Computing Assisted Delay-Aware Anomaly Detect...

引用

International Conference on Mobile Ad-hoc and Sensor Networks, MSN

作者： Chao Sang Jianhua Li Jun Wu Wu Yang School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University China Graduate School of Information Production and Systems Waseda University Japan College of Computer Science and Technology Harbin Engineering University China

ISBN: (数字)9798350358261

ISBN: (纸本)9798350358278

Today, with more and more devices in the industrial control system (ICS), the risk becomes higher and brings more attack surfaces. The need for reliable anomaly detection systems is increasing. Traditional SCADA-based detection systems deployed are difficult to assess large-scale control systems accurately, and novel AI-based technologies struggle to ensure timely response. In this paper, we propose an edge computing assisted delay-aware anomaly detection (ECADA) scheme for ICS, which considers both the accuracy and timeliness, and ensures that abnormal conditions can be accurately detected and handled in a short time. First, we model the components in ICS as three layers, taking network resources, delay, and reliability into consideration. Second, we convert the anomaly detection procedure into a decision making problem. By dividing the warning capabilities into various levels, the flexibility of the anomaly detection system is enhanced. Third, we cast a mixed-integer linear programming (MILP) problem to find the efficient anomaly detection mechanism, so that it can be dynamically scheduled to achieve the tradeoff between reliability and timeliness. We use an real-world industrial system dataset for experimental evaluation. By comparing with various traditional anomaly detection methods, it is proved that ECADA can always ensure reliable response of anomaly detection system in various network environments.

关键词： Computational modeling Industrial control Reliability engineering Sensors Mixed integer linear programming Task analysis Low latency communication

来源：评论

学校读者我要写书评

暂无评论

Knowledge System,Ontology,and Knowledge Graph of the Deep-Time Digital Earth(DDE): Progress and Perspective

引用

Journal of Earth Science 2023年第5期34卷 1323-1327页

作者： Xiumian Hu Yiwei Xu Xiaogang Ma Yunqiang Zhu Chao Ma Chao Li Hairong Lü Xinbing Wang Chenghu Zhou Chengshan Wang School of Earth Sciences and Engineering Nanjing UniversityNanjing 210023China State Key Laboratory of Palaeobiology and Stratigraphy Nanjing Institute of Geology and Palaeontology and Center for Excellence in Life and PaleoenvironmentChinese Academy of SciencesNanjing 210008China Department of Computer Science University of IdahoMoscow ID 83844USA Institute of Geographical Science and Natural Resources Chinese Academy of SciencesBeijing 100101China State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation&Institute of Sedimentary Geology&College of Computer Science and Cyber Security Chengdu University of TechnologyChengdu 610059China IUGS Deep-Time Digital Earth Research Center of Excellence Suzhou 215101China Department of Automation Tsinghua UniversityBeijing 100084China School of Electronic Information and Electrical EngineeringShanghai Jiao Tong UniversityShanghai 200240China State Key Laboratory of Biogeology and Environmental Geology China University of GeosciencesBeijing 100083China

THE USE OF KNOWLEDGE GRAPH IN NATURAL SCIENCE Knowledge graph is a field of Artificial Intelligence(AI)that aims to represent knowledge in the form of graphs,consisting of nodes and edges which represent entities and relationships between nodes respectively(Aidan et al.,2022).Although the knowledge graph was popularized recently due to use of this idea in Google’s search engine in 2012(Amit,2012),its root can be traced back to the emergence of the Semantic Web as well as earlier works in ontology(Aggarwal,2021).

关键词： DDE Google Web Semantic

来源：评论

学校读者我要写书评

暂无评论

Experimental Study on Reducing the Oscillations of a Cable-Suspended Parallel Robot for Video Capturing Purposes using Simulated Annealing and Path Planning 7

Experimental Study on Reducing the Oscillations of a Cable-S...

引用

7th International Conference on Signal Processing and Intelligent systems, ICSPIS 2021

作者： Ghanatian, Mohammad Yazdi, Mohammad Reza Hairi Masouleh, Mehdi Tale School of Mechanical Engineering University of Tehran Human and Robot Interaction Laboratory Tehran Iran School of Mechanical Engineering University of Tehran Tehran Iran School of Electrical and Computer Engineering University of Tehran Human and Robot Interaction Laboratory Tehran Iran

ISBN: (纸本)9781665409384

This paper aims to smoothen the movements of an under-constrained cable-suspended parallel robot which carries a camera for video capturing purposes, especially for video capturing of football games. This goal is achieved by means of an accurate while simple PID controller optimized by the Simulated Annealing algorithm and implemented on the joint space. Moreover, a planning strategy is considered for the joint space trajectory of the robot which guarantees the zero jerk, acceleration, and velocity at the start and the end of motions. On this regard, a septic function for each joint of the robot is considered and the corresponding boundary conditions are applied to the function to make the end-effector movements less oscillatory. This method is implemented and tested on an experimental setup while the end-effector oscillation data is recorded using an IMU sensor attached to the end-effector of the robot. Applying frequency analysis on the oscillatory data of the end-effector reveals that this simple method, on average, resulted in a 33.8% reduction in the average amplitude of the end-effector oscillations. Moreover, the maximum joint space error was decreased by 76.68% when using septic joint profile compared to the ordinary linear Cartesian trajectory planning approach. Upon applying the proposed strategy, the error of the controller has been reduced by 92.26% with respect to the previous research performed on this experimental setup. Without requiring any knowledge on the dynamic model of the robot or the natural frequencies of the end-effector or using any complex controller, this method significantly increased the smoothness and accuracy of the robot movements. The proposed method can be regarded as a definitive asset when this robot is used for video capturing purposes. © 2021 IEEE.

关键词： Simulated annealing

来源：评论

学校读者我要写书评

暂无评论

Enabling Real-time Computing and Transmission Services in Large-Scale LEO Satellite Networks

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Guo, Binquan Zhang, Zhou Atapattu, Saman Pan, Miao Yan, Ye Xiong, Zehui Li, Hongyan Xidian University State Key Laboratory of Integrated Services Networks Xi'an710071 China Tianjin300457 China Singapore University of Technology and Design Pillar of Information Systems Technology and Design 487372 Singapore Hunan University College of Computer Science and Electronic Engineering Changsha410012 China University School of Engineering MelbourneVIC3000 Australia University of Houston Department of Electrical and Computer Engineering HoustonTX77204 United States Tianjin Artificial Intelligence Innovation Center Tianjin300387 China

Enhanced by onboard computing resources and inter-satellite links (ISLs), large-scale satellite networks (SNs) can supplement terrestrial networks, facilitating real-time edge computing and transmission in remote areas at low latency. This capability benefits applications such as real-time remote object tracking, vehicular edge computing, and environmental monitoring. However, different from terrestrial vehicular networks, satellite mobility and the intermittent nature of ISLs often render many computing satellites inaccessible, necessitating efficient scheduling scheme for joint computing and communication resource allocation. To meet this requirement, we investigate joint computing and transmission strategy for real-time computing applications over SNs to reduce end-to-end delays. We formulate the real-time computing and transmission problem over time-varying SNs as an integer linear programming problem. The mathematical optimization method has an exponential time complexity of O(2|L⫟| . (4 |V⫟| + 4∣ L⫟|)2|L∣) in the worst case, which grows rapidly with the growth of network sizes. Here, |V⫟| and |L⫟| denote the number of nodes and links, respectively, within the time window {\tau }. Resorting to the designed k-shortest path-based method yields a worst-case complexity of O(|V⫟|! . |V⫟|3). While efficient under common scenarios, its running time is still relatively unstable and may become tens of times longer when computing resources become scarce. To enhance efficiency, we further design a graph-based algorithm that leverages the solution space's unique structure, thereby achieving optimal solutions with polynomial time complexity O(3|L⫟| + (2\log |V⫟|+1) |V⫟|). Extensive evaluations conducted on the world's largest Starlink system validate the efficacy of the designed methods and highlight their potential synergy. © 1967-2012 IEEE.

关键词： Satellite links

来源：评论

学校读者我要写书评

暂无评论

Coherent and incoherent laser pump on a five-level atom in a strongly coupled cavity-QED system

引用

Physical Review A 2022年第1期105卷 013701-013701页

作者： Asha Devi Sarath D. Gunapala Malin Premaratne Advanced Computing and Simulation Laboratory Department of Electrical and Computer Systems Engineering Monash University Clayton Victoria 3800 Australia Jet Propulsion Laboratory California Institute of Technology Pasadena California 91109 USA

We propose a cavity quantum electrodynamic system consisting of a five-level atom coupled to a single mode of the cavity electromagnetic field. The study is focused on the regime of strong coupling between the cavity and atom. Pump laser fields and cavity fields connect the split energy levels of the atom. Instead of the well-known two-level Dicke model obtained by adiabatic elimination of the high-energy levels, we consider the pump lasers' detunings to the atomic transitions to be very small such that we can examine the influence of the higher-energy states. We have studied the effect of an external coherent drive and incoherent pumping on these higher-energy levels and observed the enhancement of intracavity photon numbers due to quantum coherence effects. The amplification of intracavity photons is achieved even without a population inversion. However, the effect of the coherent and incoherent drive is negligible for very large detunings when the higher-energy states are adiabatically eliminated. At zero and small detunings, the system reaches the steady state at an earlier instant of time for higher incoherent pumping. We find an almost agreeable steady-state behavior of the system's exact full quantum dynamics model and its semiclassical approximation. Our model tries to accurately simulate the open system by considering the cavity decay, spontaneous decay, and dephasing of the system.

关键词： Atoms, ions, & molecules in cavities Cavity quantum electrodynamics Nanophotonics

来源：评论

学校读者我要写书评

暂无评论

Train Faster, Perform Better: Modular Adaptive Training in Over-Parameterized Models

arXiv

引用

arXiv 2024年

作者： Shi, Yubin Chen, Yixuan Dong, Mingzhi Yang, Xiaochen Li, Dongsheng Wang, Yujiang Dick, Robert Lv, Qin Zhao, Yingying Yang, Fan Lu, Tun Gu, Ning Shang, Li China and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Mathematics Statistics University of Glasgow United Kingdom Microsoft Research Asia Shanghai China Department of Engineering Science University of Oxford United Kingdom Department of Electrical Engineering and Computer Science University of Michigan United States Department of Computer Science University of Colorado Boulder United States School of Microelectronics Fudan University China

Despite their prevalence in deep-learning communities, over-parameterized models convey high demands of computational costs for proper training. This work studies the fine-grained, modular-level learning dynamics of over-parameterized models to attain a more efficient and fruitful training strategy. Empirical evidence reveals that when scaling down into network modules, such as heads in self-attention models, we can observe varying learning patterns implicitly associated with each module's trainability. To describe such modular-level learning capabilities, we introduce a novel concept dubbed modular neural tangent kernel (mNTK), and we demonstrate that the quality of a module's learning is tightly associated with its mNTK's principal eigenvalue λmax. A large λmax indicates that the module learns features with better convergence, while those miniature ones may impact generalization negatively. Inspired by the discovery, we propose a novel training strategy termed Modular Adaptive Training (MAT) to update those modules with their λmax exceeding a dynamic threshold selectively, concentrating the model on learning common features and ignoring those superfluous ones. Unlike most existing training schemes with a complete BP cycle across all network modules, MAT can significantly save computations by its partially-updating strategy and can further improve performance. Experiments show that MAT nearly halves the computational cost of model training and outperforms the accuracy of baselines. © 2024, CC BY.

关键词： Eigenvalues and eigenfunctions

来源：评论

学校读者我要写书评

暂无评论

An Optimized Ensemble Model for Prediction the Bandwidth of Metamaterial Antenna

引用

computers, Materials & Continua 2022年第4期71卷 199-213页

作者： Abdelhameed Ibrahim Hattan F.Abutarboush Ali Wagdy Mohamed Mohamad Fouad El-Sayed M.El-kenawy Computer Engineering and Control Systems Department Faculty of EngineeringMansoura UniversityMansoura35516Egypt Electrical Engineering Department College of EngineeringTaibah UniversityMedina42353Saudi Arabia Operations Research Department Faculty of Graduate Studies for Statistical ResearchCairo UniversityGiza12613Egypt Wireless Intelligent Networks Center(WINC) School of Engineering and Applied SciencesNile UniversityGizaEgypt Department of Communications and Electronics Delta Higher Institute of Engineering and TechnologyMansoura35111Egypt Faculty of Artificial Intelligence Delta University for Science and TechnologyEgypt

Metamaterial Antenna is a special class of antennas that uses metamaterial to enhance their *** size affects the quality factor and the radiation loss of the *** antennas can overcome the limitation of bandwidth for small *** learning(ML)model is recently applied to predict antenna *** can be used as an alternative approach to the trial-and-error process of finding proper parameters of the simulated *** accuracy of the prediction depends mainly on the selected *** models combine two or more base models to produce a better-enhanced *** this paper,a weighted average ensemble model is proposed to predict the bandwidth of the Metamaterial *** base models are used namely:Multilayer Perceptron(MLP)and Support Vector Machines(SVM).To calculate the weights for each model,an optimization algorithm is used to find the optimal weights of the *** Group-Based Cooperative Optimizer(DGCO)is employed to search for optimal weight for the base *** proposed model is compared with three based models and the average ensemble *** results show that the proposed model is better than other models and can predict antenna bandwidth efficiently.

关键词： Metamaterial antenna machine learning ensemble model multilayer perceptron support vector machines

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：