检索结果-内蒙古大学图书馆

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Shibuya, Tatsukichi Inoue, Nakamasa Kawakami, Rei Sato, Ikuro Department of Computer Science Institute of Science Tokyo Tokyo Japan Department of Systems and Control Engineering Institute of Science Tokyo Tokyo Japan Denso IT Laboratory Tokyo Japan

ISBN: (纸本)9798350368741

For deploying deep neural networks on edge devices with limited resources, binary neural networks (BNNs) have attracted significant attention, due to their computational and memory efficiency. However, once a neural network is binarized, finetuning it on edge devices becomes challenging because most conventional training algorithms for BNNs are designed for use on centralized servers and require storing real-valued parameters during training. To address this limitation, this paper introduces binary stochastic flip optimization (BinSFO), a novel training algorithm for BNNs. BinSFO employs a parameter update rule based on Boolean operations, eliminating the need to store real-valued parameters and thereby reducing memory requirements and computational overhead. In experiments, we demonstrated the effectiveness and memory efficiency of BinSFO in finetuning scenarios on six image classification datasets. BinSFO performed comparably to conventional training algorithms with a 70.7% smaller memory requirement. Code is released at https://***/TatsukichiShibuya/ICASSP2025_BinSFO. © 2025 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Modelling information warfare dynamics to counter propaganda using a nonlinear differential equation with a PINN-based learning approach

引用

International Journal of Information Technology (Singapore) 2024年第3期16卷 1527-1538页

作者： Pandey, Rashmikiran Pandey, Mrinal Nazarov, Alexey Nikolaevich Department of Radio Engineering and Cybernetics Moscow Institute of Physics and Technology Moscow Russian Federation Federal Research Center Computer Science and Control Russian Academy of Science Moscow Russian Federation

The widespread dissemination of misinformation and propaganda has become a crucial issue in societal conversations. This study presents an innovative framework to counter propaganda within information warfare using a modified logistic differential equation integrated with a Physics-Informed Neural Network (PINN). Employing the LIAR dataset, which reflects real-world political discourse, our model captures the intricate dynamics of propaganda. The novel approach effectively estimates underlying dynamics and unknown parameters, yielding superior accuracy (97.32%) compared to benchmark models. A parameter sensitivity analysis reveals the significant impact of the spread rate parameter, shedding light on aggressive propaganda dissemination. This research paves the way for robust methods in combating propaganda, demonstrating the effectiveness and superiority of the proposed PINN-based model over established benchmark models. © 2023, The Author(s), under exclusive licence to Bharati Vidyapeeth's institute of computer Applications and Management.

关键词： Information warfare Nonlinear differential equations PINN Propaganda Random forest classifier TF-IDF

来源：评论

学校读者我要写书评

暂无评论

A blockchain based privacy-preserving federated learning scheme for Internet of Vehicles

引用

Digital Communications and Networks 2024年第1期10卷 126-134页

作者： Naiyu Wang Wenti Yang Xiaodong Wang Longfei Wu Zhitao Guan Xiaojiang Du Mohsen Guizani School of Control and Computer Engineering North China Electric Power UniversityChina Department of Mathematics&Computer Science Fayetteville State UniversityFayettevilleNCUSA Department of Electrical and Computer Engineering Stevens Institute of TechnologyHobokenNJ07030USA Department of Computer Science and Engineering Qatar UniversityDohaQatar

The application of artificial intelligence technology in Internet of Vehicles(lov)has attracted great research interests with the goal of enabling smart transportation and traffic ***,concerns have been raised over the security and privacy of the tons of traffic and vehicle *** this regard,Federated Learning(FL)with privacy protection features is considered a highly promising ***,in the FL process,the server side may take advantage of its dominant role in model aggregation to steal sensitive information of users,while the client side may also upload malicious data to compromise the training of the global *** existing privacy-preserving FL schemes in IoV fail to deal with threats from both of these two sides at the same *** this paper,we propose a Blockchain based Privacy-preserving Federated Learning scheme named BPFL,which uses blockchain as the underlying distributed framework of *** improve the Multi-Krum technology and combine it with the homomorphic encryption to achieve ciphertext-level model aggregation and model filtering,which can enable the verifiability of the local models while achieving ***,we develop a reputation-based incentive mechanism to encourage users in IoV to actively participate in the federated learning and to practice *** security analysis and performance evaluations are conducted to show that the proposed scheme can meet the security requirements and improve the performance of the FL model.

关键词： Federated learning Blockchain Privacy-preservation Homomorphic encryption Internetof vehicles

来源：评论

学校读者我要写书评

暂无评论

An embedding observer for nonlinear dynamical systems with global convergence

An embedding observer for nonlinear dynamical systems with g...

引用

93rd Annual Meeting of the International Association of Applied Mathematics and Mechanics, GAMM 2023

作者： Gerbet, Daniel Röbenack, Klaus Institute of Control Theory Faculty of Electrical and Computer Engineering TUD Dresden University of Technology Dresden Germany

State observers for nonlinear systems are often designed for a canonical form of this system. However, this form may possess singular points, where the vector field is not defined or a Lipschitz condition is not fulfilled. This unpleasanbehavior can possibly be avoided using an embedding into a higher dimensionaspace. A construction of such an embedding and the corresponding inverse map is discussed for polynomial systems using methods from algebraic geometry. © 2023 The Authors. Proceedings in Applied Mathematics & Mechanics published by Wiley-VCH GmbH.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

On functional observers for polynomial systems with nonlinear oscillations

On functional observers for polynomial systems with nonlinea...

引用

93rd Annual Meeting of the International Association of Applied Mathematics and Mechanics, GAMM 2023

作者： Röbenack, Klaus Gerbet, Daniel Institute of Control Theory Faculty of Electrical and Computer Engineering TUD Dresden University of Technology Dresden Germany

In this contribution we discuss the design of functional observers for polynomial systems. Our approach is based on a high gain design employing an embedded observer. The functional to be estimated is generated from the high gain observer’s state. The restriction to polynomial systems allows the representation of the system’s and the observer’s dynamics as polynomial ideals. For the calculations we utilize methods from algebraic geometry. Our approach is illustrated on a chaotic system. © 2023 The Authors. Proceedings in Applied Mathematics and Mechanics published by Wiley-VCH GmbH.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adaptive Prescribed-Time Neural control of Nonlinear Systems via Dynamic Surface Technique

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第10期5卷 4948-4958页

作者： Wang, Ping Yu, Chengpu Lv, Maolong Zhao, Zilong Beijing Institute of Technology School of Automation Beijing100081 China Beijing Institute of Technology Chongqing Innovation Center Chongqing401120 China Air Force Engineering University College of Air Traffic Control and Navigation Xi'an710051 China Delft University of Technology Faculty of Electrical Engineering Mathematics and Computer Science Delft2628 CD Netherlands

The adaptive practical prescribed-time (PPT) neural control is studied for multiinput multioutput (MIMO) nonlinear systems with unknown nonlinear functions and unknown input gain matrices. Unlike existing PPT design schemes based on backstepping, this study proposes a novel PPT control framework using the dynamic surface control (DSC) approach. First, a novel nonlinear filter (NLF) with an adaptive parameter estimator and a piecewise function is constructed to effectively compensate for filter errors and facilitate prescribed-time convergence. Based on this, a unified DSC-based adaptive PPT control algorithm, augmented with a neural networks (NNs) approximator, is developed, where NNs are used to approximate unknown nonlinear system functions. This algorithm not only addresses the inherent computational complexity explosion associated with traditional backstepping methods but also reduces the constraints on filter design parameters compared to the DSC algorithm that relies on linear filters. The simulation showcases the effectiveness and superiority of the devised scheme by employing a two-degree-of-freedom robot manipulator. © 2024 IEEE.

关键词： Flexible manipulators

来源：评论

学校读者我要写书评

暂无评论

DialogNTM: Context Reconstruction in Multi-Turn Dialogue Generation using Neural Turing Machines

DialogNTM: Context Reconstruction in Multi-Turn Dialogue Gen...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Zhao, Haohao Liu, Meiling Zhou, Jiyun Fu, Kaiquan College of Computer and Control Engineering Northeast Forestry University Harbin China Lieber Institute Johns Hopkins University Baltimore United States South Dakota State University Department of Electrical Engineering and Computer Science BrookingsSD United States

ISBN: (纸本)9781665410205

In actual conversational scenarios, we can often determine which parts of the previous dialogue are more critical based on the current inquiry. However, the existing contextual modeling methods often encode the query sentence and the dialogue history in a unified manner, which fails to effectively highlight the inference effect of the query sentence. Moreover, these methods typically process the dialogue history only at the information extraction level, neglecting the treatment of the context itself. In this paper, we propose a novel conversational context modeling technique called DialogNTM. Based on the guidance of the query sentence, the technology can effectively eliminate redundant information by reconstructing the representation of the context. Specifically, we have tweaked the memory and input flow of the Neural Turing Machine (NTM) to encode contextual information in memory and guide the read, write, and erase operations of memory through query sentence. This design simulates the human brain's dynamic retrieval and renewal mechanism of previous memories when dealing with current problems. We have conducted extensive experiments on three publicly available datasets to verify the effectiveness of the DialogNTM model. Compared to the benchmark model, DialogNTM showed significant performance improvements ranging from 11% to 73% across multiple automated evaluation metrics (3.52% to 8.68% in absolute terms). © 2024 IEEE.

关键词： Turing machines

来源：评论

学校读者我要写书评

暂无评论

Green and Low Carbon Load Side Carbon Emission Measurement in Power Systems Based on Opportunity Constraints 11

Green and Low Carbon Load Side Carbon Emission Measurement i...

引用

11th China International Conference on Electricity Distribution, CICED 2024

作者： Xu, Zhiyuan Cheng, Liang School of Control and Computer Engineering North China Electric Power University Beijing China Institute of Software Chinese Academy of Sciences Tca Lab Beijing China

ISBN: (纸本)9798350368345

In the context of global warming, traditional carbon measurement methods have failed to completely capture the carbon emission dynamics of the power system under various operating states, resulting in biased carbon emission data. Moreover, there is still significant research space for measurement models on the power load side. To improve the accuracy and real-time performance of carbon emission measurement, this study designs a power system load side carbon emission measurement model based on opportunity constraints. The proposed model is built on opportunity constraints and fairness principles to allocate carbon emission responsibilities on the load side. On this basis, considering the dynamic characteristics of the power system, a two-stage carbon emission measurement model is raised, which uses dynamic carbon emission measurement method and node carbon potential method to calculate carbon emissions in the first and second stages, respectively. The outcomes validated that the carbon emissions of Unit 1 in the fixed carbon emission factor model were 620.07 tons, while they were reduced to 470.11 tons in the proposed model. After implementing demand response measures, the total carbon emissions of the system during peak electricity usage decreased by 3130 kg. The fixed carbon emission factor econometric model, traditional carbon market model, and the proposed model resulted in a total carbon emissions of 18306.61 tons, 16939.79 tons, and 19045.21 tons, respectively. Therefore, the proposed model performs better in regulating demand, can more accurately measure carbon emissions on the load side of the power system, and contribute to promoting green and low-carbon development of the energy system. © 2024 IEEE.

关键词： Low emission

来源：评论

学校读者我要写书评

暂无评论

Research on Autonomous Decision-Making of Multi-UAV Air Combat Based on Deep Reinforcement Learning 7th

Research on Autonomous Decision-Making of Multi-UAV Air Comb...

引用

7th Chinese Conference on Swarm Intelligence and Cooperative control, CCSICC 2023

作者： Zhou, Zhiming Qi, Hongmin Liu, Zhen Qian, Dianwei Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Control and Computer Engineering North China Electric Power University Beijing102206 China

ISBN: (纸本)9789819733354

To solve the autonomous decision-making problem of multi-UAV air combat, a multi-UAV reinforcement learning self-game autonomous decision-making method for air combat with parameter sharing is proposed. Firstly, a self-game policy pool training framework is designed. The game ability of the multi-UAV autonomous decision-making model is improved by self-game confrontation with a policy pool containing different training generations. Secondly, a parameter-sharing mechanism is proposed to train homogeneous UAVs to improve the efficiency of reinforcement learning training and the utilization of interactive data. Finally, a multi-UAV autonomous decision-making model using the multi-agent deep deterministic policy gradient algorithm is designed so that it can explore effectively in the continuous action space. In addition, experiments are designed to verify the effectiveness and migration of the designed algorithm and training framework. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

A Comparative Evaluation of GA and PSO Tuned PID controllers for the Quadruple Tank System

A Comparative Evaluation of GA and PSO Tuned PID Controllers...

引用

2024 International Conference on Information and Communication Technology for Development for Africa, ICT4DA 2024

作者： Hailu, Gashaye L. Yetayew, Tefera T. Bahir Dar University Electrical and Computer Engineering Faculty Bahir Dar Institute of Technology Bahir Dar Ethiopia Adama Science and Technology University Department of Power and Control Engineering Adama Ethiopia

ISBN: (纸本)9798331527792

In this research, nature inspired metaheuristic optimization algorithms: Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) Techniques are formulated to tune optimal combinations of PID controller parameters for a quadruple tank level control application. The degree of relation of the best paring for both interacted control loops are justified by using Relative Gain Array (RGA) calculation. The overall system has been modeled and implemented in MATLAB/Simulink and performance evaluations has been done for time domain performance specifications of settling time, rise time, percentage overshoot and steady state error as comparison criterions of controllers. Accordingly, for the minimum phase case, GA tuned PID has settling time of 0.36sec and 0.31second for the tank 1 and tank 2 respectively. And PSO tuned PID controller has settling time of 0.96sec and 0.35sec for tank 1and tank 2 respectively. The quantitative and qualitative analysis of results reveals that there is fast dynamic response using GA PID relative to PSO PID even though the difference is not significant that may be due to stochastic feature of the metahueristic optimization algorithms. In summary, the optimal combination of PID controller algorithms' parameters can give robust performance compared to PID controller designed using the classical approaches or manual tuning techniques. © 2024 IEEE.

关键词： Genetic algorithms

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：