检索结果-南通市图书馆

Multi-party privacy-preserving decision tree training with a privileged party

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第8期67卷 253-268页

作者： Yiwen tong Qi FENG Min LUO Debiao HE Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan University Shanghai Technology Innovation Centre of Distributed Privacy-Preserving Artificial Intelligence Matrix Elements Technologies Key Laboratory of Computing Power Network and Information Security Ministry of EducationShandong Computer Science Center Qilu University of Technology(Shandong Academy of Sciences)

Currently, a decision tree is the most commonly used data mining algorithm for classification tasks. While a significant number of studies have investigated privacy-preserving decision trees, the methods proposed in these studies often have shortcomings in terms of data privacy breach or efficiency. Additionally,these methods typically only apply to symmetric frameworks, which consist of two or more parties with equal privilege, and are not suitable for asymmetric scenarios where parties have unequal privilege. In this paper,we propose SecureCART, a three-party privacy-preserving decision tree training scheme with a privileged party. We adopt the existing pMPL framework and design novel secure interactive protocols for division,comparison, and asymmetric multiplication. Compared to similar schemes, our division protocol is 93.5–560.4× faster, with the communication overhead reduced by over 90%; further, our multiplication protocol is approximately 1.5× faster, with the communication overhead reduced by around 20%. Our comparison protocol based on function secret sharing maintains good performance when adapted to pMPL. Based on the proposed secure protocols, we implement SecureCART in C++ and analyze its performance using three real-world datasets in both LAN and WAN environments. he experimental results indicate that SecureCART is significantly faster than similar schemes proposed in past studies, and that the loss of accuracy while using SecureCART remains within an acceptable range.

关键词： decision trees privacy protection privileged party secret sharing secure multi-party computation

同方期刊数据库

Investigation of Eccentric PM on High-frequency Vibration in FSCW PM Machine Considering Force Modulation Effect

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electrical Engineering 2024年第2期10卷 59-69页

作者： Jinghua Ji Weizhi Jiang Wenxiang Zhao tong Liu School of Electrical and Information Engineering Jiangsu UniversityZhenjiang 212013China

This study investigates the negative influence of an eccentric permanent-magnet(PM)design on high-frequency electromagnetic vibration in fractional-slot concentrated-winding(FSCW)PM ***,an analytical expression for the sideband current harmonics was derived using the double Fourier series expansion ***,the characteristics of the flux-density harmonics are studied from the perspective of the space-time distribution and initial phase *** influence of the eccentric PM design on high-frequency electromagnetic and concentrated forces was studied based on the electromagnetic force modulation ***,an eccentric PM design is not conducive to reducing the 2pth-order high-frequency electromagnetic ***,two FSCW PM machines with conventional and eccentric PM designs are manufactured and tested to verify the theoretical *** results show that the eccentric PM design worsens high-frequency vibrations.

关键词： Eccentric permanent magnet machine sideband current harmonics electromagnetic force modulation effect fractional-slot concentrated-winding high-frequency vibration

Image attention transformer network for indoor 3D object detection

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(Technological Sciences) 2024年第7期67卷 2176-2190页

作者： REN KeYan YAN tong HU ZhaoXin HAN HongGui ZHANG YunLu Faculty of Information Technology Beijing University of TechnologyBeijing 100124China

Point clouds and RGB images are both critical data for 3D object detection. While recent multi-modal methods combine them directly and show remarkable performances, they ignore the distinct forms of these two types of data. For mitigating the influence of this intrinsic difference on performance, we propose a novel but effective fusion model named LI-Attention model, which takes both RGB features and point cloud features into consideration and assigns a weight to each RGB feature by attention ***, based on the LI-Attention model, we propose a 3D object detection method called image attention transformer network(IAT-Net) specialized for indoor RGB-D scene. Compared with previous work on multi-modal detection, IAT-Net fuses elaborate RGB features from 2D detection results with point cloud features in attention mechanism, meanwhile generates and refines 3D detection results with transformer model. Extensive experiments demonstrate that our approach outperforms stateof-the-art performance on two widely used benchmarks of indoor 3D object detection, SUN RGB-D and NYU Depth V2, while ablation studies have been provided to analyze the effect of each module. And the source code for the proposed IAT-Net is publicly available at https://***/wisper181/IAT-Net.

关键词： 3D object detection transformer attention mechanism

维普期刊数据库

同方期刊数据库

Lightweight Si_(3)N_(4)@SiO_(2) ceramic foam for thermal insulation and electromagnetic wave transparency

在线全文

学校读者我要写书评

暂无评论

Nano Research 2024年第5期17卷 4298-4306页

作者： Zongwei tong Xiangjie Yan Yuexiang Wang Kexun Li Weihai Ma China-Blarus Belt and Road Joint Laboratory on Electromagnetic Environment Effect the 33rd Research Institute of ChinaElectronics Technology Group CorporationTaiyuan 030032China Key Laboratory for Advanced Ceramics and Machining Technology of Ministry of Education School of Material Science and EngineeringTianjin UniversityTianjin 300072China Department of Materials Science and Engineering Jinzhong UniversityJinzhong 030619China

Increasing porosity is one of the most direct ways to improve the thermal insulation and dielectric properties of *** now,many wet methods for preparing Si_(3)N_(4) ceramic foams usually face the problems of complex rheology,long period,and expensive cost,and the reported pore sizes of Si_(3)N_(4) ceramic foams are typically micron-grade,resulting in a lack of competitiveness in thermal insulation and wave-transparent *** this paper,the Si_(3)N_(4)@SiO_(2) ceramic foams were prepared using an efficient dry-method,which combined three processes of low temperature chemical vapor deposition(LTCVD),template,and isostatic *** method has the advantages of simple operation and short preparation period,and can realize near-net size molding and mass *** addition,the evolution mechanisms of honeycomb microstructure and composition of Si_(3)N_(4)@SiO_(2) ceramic foam during sintering were studied by chemical reaction *** as-prepared Si_(3)N_(4)@SiO_(2) ceramic foam possesses low density(0.377 g·cm^(-3)),high compressive strength(7.5 MPa),low thermal conductivity(0.0808 W·m^(-1)·K^(-1)),and excellent dielectric properties(ε<1.32,tanδ<0.009)in the frequency range of 8-18 GHz,and its maximum working temperature in air can reach up to 1100℃.It will be recommended to be applied in the interlayer of Si_(3)N_(4) ceramic radome to improve its thermal insulation and electromagnetic wave transparency performances.

关键词： silicon nitride ceramic foam honeycomb microstructure dielectric properties thermal insulation

The source of Double ITCZ induced by the SST bias over the tropical western Pacific as reflected in CAS-ESM2 Model

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 Earth Sciences 2024年第5期67卷 1604-1615页

作者： Jinning tong Fei ZHENG Jiangbo JIN Ruowen YANG Jiang ZHU Qingcun ZENG Department of Atmospheric Sciences Yunnan UniversityKunming 650091China International Center for Climate and Environment Science Institute of Atmospheric PhysicsChinese Academy of SciencesBeijing 100029China

“Double ITCZ”is a common precipitation bias over the tropical Pacific in current climate models and Earth system models,but the reasons for its formation are still worth exploring and *** this study,we adopted the second version of Chinese Academy of Sciences Earth System Model(CAS-ESM2),by comparing a set of sea surface temperature(SST)bias correction experiments over the tropical Pacific,to explore the possible mechanism of SST bias on inducing the“Double ITCZ”from the perspective of the climatic annual mean bias in the coupled *** revealed that the simulated climatic annual mean SST bias over the tropical Pacific can affect the bias of latent heat flux through the saturation specific humidity,and the bias of latent heat flux can further affect that of vertical velocity of humid air by the condensation release mechanism,and finally modulate the simulated bias in ***,through the inter-comparison between different experiments,it is found that the source of Double ITCZ might mainly come from the annual mean SST bias over the tropical western Pacific through the proposed air-sea coupled process of“SST-saturated specific humidity-latent heat flux-vertical velocity-precipitation”,indicating a possible way on reducing the couple biases in models over the tropical Pacific to improve the accuracy of CAS-ESM2 for climate simulation.

关键词： SST bias Double ITCZ Nudging SST CAS-ESM2

维普期刊数据库

Impact of scandium and terbium on the mechanical properties,corrosion behavior,and biocompatibility of biodegradable Mg-Zn-Zr-Mn alloys

在线全文

学校读者我要写书评

暂无评论

Journal of Magnesium and Alloys 2024年第2期12卷 546-572页

作者： Khurram Munir Jixing Lin Xian tong Arne Biesiekierski Yuncang Li Cuie Wen School of Engineering RMIT UniversityMelbourneVictoria 3001Australia School and Hospital of Stomatology Wenzhou Medical UniversityWenzhou 325027China

Magnesium(Mg)-based bone implants degrade rapidly in the physiological environment of the human body which affects their structural integrity and biocompatibility before adequate bone *** earth elements(REEs)have demonstrated their effectiveness in tailoring the corrosion and mechanical behavior of Mg *** study methodically investigated the impacts of scandium(Sc)and terbium(Tb)in tailoring the corrosion resistance,mechanical properties,and biocompatibility of Mg–0.5Zn–0.35Zr–0.15Mn(MZZM)alloys fabricated via casting and hot *** indicate that addition of Sc and Tb improved the strength of MZZM alloys via grain size reduction and solid solution strengthening *** extruded MZZM–(1–2)Sc–(1–2)Tb(wt.%)alloys exhibit compressive strengths within the range of 336–405 MPa,surpassing the minimum required strength of 200 MPa for bone implants by a significant *** polarization tests revealed low corrosion rates of as–cast MZZM(0.25 mm/y),MZZM–2Tb(0.45 mm/y),MZZM–1Sc–1Tb(0.18 mm/y),and MZZM–1Sc–2Tb(0.64 mm/y),and extruded MZZM(0.17 mm/y),MZZM–1Sc(0.15 mm/y),MZZM-2Sc(0.45 mm/y),MZZM-1Tb(0.17 mm/y),MZZM-2Tb(0.10 mm/y),MZZM–1Sc-1Tb(0.14 mm/y),MZZM-1Sc-2Tb(0.40 mm/y),and MZZM–2Sc–2Tb(0.51 mm/y)alloys,which were found lower compared to corrosion rate of high-purity Mg(~1.0 mm/y)reported in the ***,addition of Sc,or Tb,or Sc and Tb to MZZM alloys did not adversely affect the viability of SaOS2 cells,but enhanced their initial cell attachment,proliferation,and spreading shown via polygonal shapes and *** study emphasizes the benefits of incorporating Sc and Tb elements in MZZM alloys,as they effectively enhance corrosion resistance,mechanical properties,and biocompatibility simultaneously.

关键词： Corrosion property In vitro cytotoxicity Magnesium alloys Mechanical property Rare earth elements

维普期刊数据库

Toward a Direct Measurement of the Cosmic Acceleration: The Pilot Observation of HI 21 cm Absorption Line at FAST

在线全文

学校读者我要写书评

暂无评论

Research in Astronomy and Astrophysics 2024年第7期24卷 50-58页

作者： Jiangang Kang Chang-Zhi Lu tong-Jie Zhang Ming Zhu Institute for Frontiers in Astronomy and Astrophysics Beijing Normal University School of Physics and Astronomy Beijing Normal University National Astronomical Observatories Chinese Academy of Sciences

This study presents results on detecting neutral atomic hydrogen（H I） 21 cm absorption in the spectrum of PKS PKS1413+13 at redshift z = 0.24670041. The observation was conducted by FAST, with a spectral resolution of10 Hz, using 10 minutes of observing time. The global spectral profile is examined by modeling the absorption line using a single Gaussian function with a resolution of 10 kHz within a 2 MHz bandwidth. The goal is to determine the rate of the latest cosmic acceleration by directly measuring the redshift evolution of the H I 21 cm absorption line with Hubble flow toward a common background quasar over a decade or longer time span. This will serve as a detectable signal generated by the accelerated expansion of the Universe at redshift z < 1, referred to as redshift drift z（5） or the SL effect. The measured H I gas column density in this DLA system is approximately equivalent to the initial observation value, considering uncertainties of the spin temperature of a spiral host galaxy. The high signal-to-noise ratio of 57,obtained at a 10 kHz resolution, strongly supports the feasibility of using the H I 21 cm absorption line in DLA systems to accurately measure the redshift drift rate at a precision level of around 10-10per decade.

关键词： (cosmology:) cosmological parameters (cosmology:) dark energy cosmology: observations cosmology: theory ISM: clouds radio lines: galaxies radio lines: ISM

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

Aspect-Level Sentiment Analysis Based on Deep Learning

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第3期78卷 3743-3762页

作者： Mengqi Zhang Jiazhao Chai Jianxiang Cao Jialing Ji tong Yi Department of Communication University of New South WalesSydney2052Australia Higher Vocational Education Publishing Division Higher Education PressBeijing100029China State Key Laboratory of Media Convergence and Communication Communication University of ChinaBeijing100020China Key Lab of Education Blockchain and Intelligent Technology Ministry of EducationGuangxi Normal UniversityGuilin541004China

In recent years,deep learning methods have developed rapidly and found application in many fields,including natural language *** the field of aspect-level sentiment analysis,deep learning methods can also greatly improve the performance of ***,previous studies did not take into account the relationship between user feature extraction and contextual *** address this issue,we use data feature extraction and deep learning combined to develop an aspect-level sentiment analysis *** be specific,we design user comment feature extraction(UCFE)to distill salient features from users’historical comments and transform them into representative user feature ***,the aspect-sentence graph convolutional neural network(ASGCN)is used to incorporate innovative techniques for calculating adjacency matrices;meanwhile,ASGCN emphasizes capturing nuanced semantics within relationships among aspect words and syntactic dependency ***,three embedding methods are devised to embed the user feature vector into the ASGCN *** empirical validations verify the effectiveness of these models,consistently surpassing conventional benchmarks and reaffirming the indispensable role of deep learning in advancing sentiment analysis methodologies.

关键词： Aspect-level sentiment analysis deep learning graph convolutional neural network user features syntactic dependency tree