检索结果-南通市图书馆

Heterogeneous decentralised machine unlearning with seed model distillation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

CAAI Transactions on Intelligence Technology 2024年第3期9卷 608-619页

作者： Guanhua Ye Tong Chen quoc viet hung nguyen Hongzhi Yin School of Information Technology and Electrical Engineering The University of QueenslandSt LuciaQueenslandAustralia School of Information and Communication Technology Griffith UniversityBrisbaneQueenslandAustralia

As some recent information security legislation endowed users with unconditional rights to be forgotten by any trained machine learning model,personalised IoT service pro-viders have to put unlearning functionality into their *** most straight-forward method to unlearn users'contribution is to retrain the model from the initial state,which is not realistic in high throughput applications with frequent unlearning *** some machine unlearning frameworks have been proposed to speed up the retraining process,they fail to match decentralised learning scenarios.A decentralised unlearning framework called heterogeneous decentralised unlearning framework with seed(HDUS)is designed,which uses distilled seed models to construct erasable en-sembles for all ***,the framework is compatible with heterogeneous on-device models,representing stronger scalability in real-world *** experiments on three real-world datasets show that our HDUS achieves state-of-the-art performance.

关键词： data mining data privacy machine learning

Privacy-preserving explainable AI: a survey

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年

As the adoption of explainable AI (XAI) continues to expand, the urgency to address its privacy implications intensifies. Despite a growing corpus of research in AI privacy and explainability, there is little attention on privacy-preserving model explanations. This article presents the first thorough survey about privacy attacks on model explanations and their countermeasures. Our contribution to this field comprises a thorough analysis of research papers with a connected taxonomy that facilitates the categorization of privacy attacks and countermeasures based on the targeted explanations. This work also includes an initial investigation into the causes of privacy leaks. Finally, we discuss unresolved issues and prospective research directions uncovered in our analysis. This survey aims to be a valuable resource for the research community and offers clear insights for those new to this domain. To support ongoing research, we have established an online resource repository, which will be continuously updated with new and relevant findings.

关键词：

Landslide Susceptibility Mapping Using RBFN-Based Ensemble Machine Learning Models

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2025年第1期142卷 467-500页

作者： Landslide Susceptibility Mapping Using RBFN-Based Ensemble Machine Learning Models Duc-Dam nguyen nguyen viet Tiep Quynh-Anh Thi Bui Hiep Van Le Indra Prakash Romulus Costache Manish Pandey Binh Thai Pham Faculty of Civil Engineering University of Transport TechnologyThanh XuanHa Noi100000Vietnam Faculty of Hydraulic Engineering Hanoi University of Civil EngineeringHa Noi100000Vietnam DDG(R)Geological Survey of India Gandhinagar382010India Research Institute of the University of Bucharest Bucharest050663Romania National Institute of Hydrology and Water Management Bucharest013686Romania Department of Civil Engineering Transilvania University of BrasovBrasov500152Romania Danube Delta National Institute for Research and Development Tulcea820112Romania Marwadi University Research Center(MURC) Marwadi UniversityRajkotGujaratBharat360003India Department of Civil Engineering Faculty of Engineering&TechnologyMarwadi UniversityRajkotGujaratBharat360003India

This study was aimed to prepare landslide susceptibility maps for the Pithoragarh district in Uttarakhand,India,using advanced ensemble models that combined Radial Basis Function Networks(RBFN)with three ensemble learning techniques:DAGGING(DG),MULTIBOOST(MB),and ADABOOST(AB).This combination resulted in three distinct ensemble models:DG-RBFN,MB-RBFN,and ***,a traditional weighted method,Information Value(IV),and a benchmark machine learning(ML)model,Multilayer Perceptron Neural Network(MLP),were employed for comparison and *** models were developed using ten landslide conditioning factors,which included slope,aspect,elevation,curvature,land cover,geomorphology,overburden depth,lithology,distance to rivers and distance to *** factors were instrumental in predicting the output variable,which was the probability of landslide *** analysis of the models’performance indicated that the DG-RBFN model,with an Area Under ROC Curve(AUC)of 0.931,outperformed the other *** AB-RBFN model achieved an AUC of 0.929,the MB-RBFN model had an AUC of 0.913,and the MLP model recorded an AUC of *** results suggest that the advanced ensemble ML model DG-RBFN was more accurate than traditional statistical model,single MLP model,and other ensemble models in preparing trustworthy landslide susceptibility maps,thereby enhancing land use planning and decision-making.

关键词： Landslide susceptibility map spatial analysis ensemble modelling information values(IV)

Design and Demonstration of a Locust-Like Jumping Mechanism for Small-Scale Robots

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2012年第3期9卷 271-281页

作者： quoc-viet nguyen Hoon Cheol Park Department of Advanced Technology Fusion Konkuk University Seoul 143-701 Republic of Korea

A jumping mechanism can be an efficient mode of motion for small robots to overcome large obstacles on the ground and rough terrain. In this paper, we present a 7 g prototype of locust-inspired jumping mechanism that uses springs, wire, reduction gears, and a motor as the actuation components. The leg structure and muscles of a locust or grasshopper were mimicked using springs and wire, springs for passive extensor muscles, and a wire as a flexor muscle. A small motor was used to slowly charge the spring through a lever and gear system, and a cam with a special profile was used as a clicking mechanism for quick release of elastic energy stored in the springs to create a sudden kick for a quick jump. Performance analysis and experiments were conducted for comparison and performance estimation of the jumping mechanism prototype. Our prototype could produce standing jumps over obstacles that were about 14 times its own size （approximate to 71 cm） and a jumping distance of 20 times its own size （approximate to 100 cm）.

关键词： biomimetics locust jumping mechanism passive tibia extensor muscle

Hybrid Design and Performance Tests of a Hovering Insect-inspired Flapping-wing Micro Aerial Vehicle

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2016年第2期13卷 235-248页

作者： quoc-viet nguyen Woei Leong Chan Marco Debiasi Temasek Laboratories National University of Singapore Singapore 117411 Singapore

Hovering ability is one of the most desired features in Flapping-Wing Micro Air Vehicles （FW-MAVs）. This paper presents a hybrid design of flapping wing and fixed wing, which combines two flapping wings and two fixed wings to take advantage of the double wing clap-and-fling effect for high thrust production, and utilizes the fixed wings as the stabilizing surfaces for inherently stable hovering flight. Force measurement shows that the effect of wing clap-and-fling significantly enhances the cycle-averaged vertical thrust up to 44.82% at 12.4 Hz. The effect of ventral wing clap-and-fling due to presence of fixed wings produces about 11% increase of thrust-to-power ratio, and the insect-inspired FW-MAV can produce enough cycle-averaged vertical thrust of 14.76 g for lift-offat 10 Hz, and 24 g at maximum frequency of 12.4 Hz. Power measurement indicates that the power consumed for aerodynamic forces and wing inertia, and power loss due to gearbox friction and mechanism inertia was about 80% and 20% of the total input power, respectively. The proposed insect-inspired FW-MAV could endure three-minute flight, and demonstrate a good flight performance in terms of vertical take-off, hovering, and control with an onboard 3.7 V-70 mAh LiPo battery and control system.

关键词： insect-inspired flapping wing hovering clap-and-fling MAV vertical takeoff

Fresh and hardened properties of high-performance fiber-reinforced concrete containing fly ash and rice husk ash:Influence of fiber type and content

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Structural and Civil Engineering 2022年第12期16卷 1621-1632页

作者： nguyen-Trong HO viet quoc DANG Minh-Hieu nguyen Chao-Lung HWANG Trong-Phuoc HUYNH Faculty of Civil Engineering VSB-Technical University of OstravaOstrava-Poruba 70800Czech Republic Faculty of Bridge and Road Construction Mientrung University of Civil Engineering(MUCE)Tuy Hoa City 56100Vietnam Vinh Long Department of Construction Vinh Long City 85000Vietnam Taiwan Building Technology Center Taiwan University of Science and TechnologyTaipei 10607China Faculty of Civil Engineering College of EngineeringCan Tho UniversityCan Tho City 94000Vietnam

Although fibers are used only infrequently as an additive in concrete in the construction industry,fiberenhanced concrete is known to provide a wide range of advantages over conventional *** main objective of this study was to investigate the influences of fiber type and content on the mechanical properties and durability of highperformance fiber-reinforced concrete(HPFRC)designed using a novel densified mixture design algorithm with fly ash and rice husk *** types of fiber,including polypropylene(PP)fiber,steel fiber(SF),and hybrid fiber(HF),were *** on the results,the inclusion of fibers decreased HPFRC flowability,regardless of fiber *** the compressive strength of HPFRC with 1.6%PP fiber content was 11.2%below that of the reference HPFRC specimen at 91 d of curing age,the 91-d compressive strengths of both SF and HF-enhanced HPFRC specimens were significantly better than that of the reference HPFRC ***,the HPFRC specimens incorporating SF and HF both exhibited better splitting tensile and flexural strengths as well as less drying shrinkage than the HPFRC specimens incorporating PP ***,the fiber-enhanced specimens,especially those with added SF,registered less surface electrical resistivity and greater vulnerability to chloride ion penetration than the reference HPFRC specimen.

关键词： high-performance fiber-reinforced concrete fly ash rice husk ash durability mechanical strength

维普期刊数据库

Development of an automatic and knowledge-infused framework for structural health monitoring based on prompt engineering

在线全文

学校读者我要写书评

暂无评论

FRONTIERS OF STRUCTURAL AND CIVIL ENGINEERING 2024年第11期18卷 1752-1774页

作者： nguyen, Truong-Thang Dang, viet-hung Pham, Thanh-Tung Hanoi Univ Civil Engn Fac Bldg & Ind Construct Hanoi 11600 Vietnam Hanoi Univ Civil Engn Res Grp Dev & applicat Adv Mat & Modern Technol Co Hanoi 11600 Vietnam

Collecting and analyzing vibration signals from structures under time-varying excitations is a non-destructive structural health monitoring approach that can provide meaningful information about the structures' safety without interrupting their normal operations. This paper develops a novel framework using prompt engineering for seamlessly integrating users' domain knowledge about vibration signals with the advanced inference ability of well-trained large language models (LLMs) to accurately identify the actual states of structures. The proposed framework involves formulating collected data into a standardized form, utilizing various prompts to gain useful insights into the dynamic characteristics of vibration signals, and implementing an in-house program with the help of LLMs to perform damage detection. The advantages, as well as limitations, of the proposed method are qualitatively and quantitatively assessed through two realistic case studies from literature, demonstrating that the present method is a new way to quickly construct practical and reliable structural health monitoring applications without requiring advanced programming/mathematical skills or obscure specialized programs.

关键词： structural health monitoring vibration large language model signal processing prompt engineering

Verification of Beam Models for Ionic Polymer-Metal Composite Actuator

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2009年第3期6卷 232-238页

作者： Hoon Cheol Park quoc viet nguyen Jang Woo Lee Young Tai Yoo Artificial Muscle Research Center Konkuk University Department of Advanced Technology Fusion Konkuk University Department of Materials Chemistry and Engineering Konkuk University

Ionic Polymer-Metal Composite (IPMC) can work as an actuator by applying a few voltages.A thick IPMC actuator,where Nafion-117 membrane was synthesized with polypyrrole/alumina composite filler,was analyzed to verify the equivalent beam and equivalent bimorph beam *** blocking force and tip displacement of the IPMC actuator were measured with a DC power supply and Young's modulus of the IPMC strip was measured by bending and tensile tests *** calculated maximum tip displacement and the Young's modulus by the equivalent beam model were almost identical to the corresponding measured *** element analysis with thermal analogy technique was utilized in the equivalent bimorph beam model to numerically reproduce the force-displacement relationship of the IPMC *** results by the equivalent bimorph beam model agreed well with the force-displacement relationship acquired by the measured *** is confirmed that the equivalent beam and equivalent bimorph beam models are practically and effectively suitable for predicting the tip displacement,blocking force and Young's modulus of IPMC actuators with different thickness and different composite of ionic polymer membrane.

关键词： jonic polymer-metal composite beam model displacement force

维普期刊数据库

Forecasting measured responses of structures using temporal deep learning and dual attention

在线全文

学校读者我要写书评

暂无评论

Frontiers of Structural and Civil Engineering 2024年第6期18卷 832-850页

作者： viet-hung DANG Trong-Phu nguyen Thi-Lien PHAM Huan X.nguyen Department of Structural Mechanics Hanoi University of Civil EngineeringHanoi 11600Vietnam Faculty of Applied Science University of Economics-Technology for IndustriesHanoi 11600Vietnam London Digital Twin Research Centre Faculty of Science and TechnologyMiddlesex UniversityLondon NW44BTUK International School Vietnam National UniversityHanoi 11600Vietnam

The objective of this study is to develop a novel and efficient model for forecasting the nonlinear behavior of structures in response to time-varying random *** key idea is to design a deep learning architecture to leverage the relationships,between external excitations and structure's vibration signals,and between historical values and future values,within multiple time-series *** proposed method consists of two main steps:the first step applies a global attention mechanism to combine multiple-measured time series and time-varying excitation into a weighted time series before feeding it to a temporal architecture;the second step utilizes a self-attention mechanism followed by a fully connected layer to predict multi-step future *** viability of the proposed method is demonstrated via two case studies involving synthetic data from a three-dimensional(3D)reinforced concrete structure and experimental data from an 18-story steel ***,comparison and robustness studies are carried out,showing that the proposed method outperforms conventional methods and maintains high performance in the presence of noise with an amplitude of less than 10%.

关键词： structural dynamic time-varying excitation deep learning signal processing response forecasting