检索结果-南通市图书馆

Maximizing conditional independence for unsupervised domain adaptation

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年第5期67卷 136-149页

作者： yiming ZHAI Chuanxian REN Youwei LUO Daoqing DAI School of Mathematics Sun Yat-Sen University

Unsupervised domain adaptation(UDA) studies how to transfer a learner from a labeled source domain to an unlabeled target domain with different distributions. Existing methods mainly focus on matching marginal distributions of the source and target domains, which probably leads to a misalignment of samples from the same class but different domains. In this paper, we tackle this misalignment issue by achieving the class-conditioned transferring from a new perspective. Specifically, we propose a method named maximizing conditional independence(MCI) for UDA, which maximizes the conditional independence of feature and domain given class in the reproducing kernel Hilbert spaces. The optimization of conditional independence can be viewed as a surrogate for minimizing class-wise mutual information between feature and domain. An interpretable empirical estimation of the conditional dependence measure is deduced and connected with the unconditional case. Besides, we provide an upper bound on the target error by taking the class-conditional distribution into account, which provides a new theoretical insight for class-conditioned transferring. Extensive experiments on six benchmark datasets and various ablation studies validate the effectiveness of the proposed model in dealing with UDA.

关键词： conditional independence kernel method domain adaptation class-conditioned transferring

Characteristics of bioaerosols under high-ozone periods,haze episodes,dust storms,and normal days in Xi'an,China

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Particuology 2024年第7期90卷 140-148页

作者： yiming Yang Liu Yang Xiaoyan Hu Zhenxing Shen Department of Environmental Science and Engineering Xi’an Jiaotong UniversityXi’an710049China Key Lab of Aerosol Chemistry&Physics SKLLQGInstitute of Earth EnvironmentChinese Academy of SciencesXi’an710049China

The characteristics of bioaerosols and their effects on human beings'health have become a major public concern in the recent *** study compared the characteristics of bioaerosols under different types of pollution(high-ozone[HO]episodes,haze episodes,and dust storms[DSs]),with those of bioaerosols on normal days(NDs)in Xi'an,*** concentrations,diurnal variations,and size distribution of bacteria,fungi,and total airborne microbes(TAMs)were *** results showed that the bacterial and fungal concentrations were the highest during DSs and on NDs,reaching 42..2.and 956.9 colony-forming units/m^(3),*** concentration of TAMs was the highest during HO episodes,reaching 10.7×10^(6)cells/m^(3),which was significantly higher than that during the other *** maximum concentrations of bacteria and fungi during HO episodes were observed in the morning,while they appeared separately in the morning and evening during haze *** and TAMs during DSs reached peak concentrations in the morning and evening,which is opposite to the particulate matter(PM)distribution *** NDs,temperature and relative humidity had significant effects on *** size distributions of airborne bacteria exhibited bimodal patterns,whereas unimodal distributions were observed for *** mainly attached in fine particles(2.1μm)with proportions of 67.4%and 60.5%during haze episodes and DSs,respectively,which means that fungi were easily penetrated into the human respiratory *** evaluation of influence factors revealed that PM_(2.5) and PM10 had negative effects on fungal concentration during DSs,and ozone concentration and temperature were inversely correlated with fungal concentration during HO *** results provide valuable reference data for elucidating the formation and evolution of bioaerosols under different types of pollution.

关键词： Bacteria Fungi Total airborne microbe Different air pollution types Influence factors

Plastic deformation behavior of a Cu-10Ta alloy under strong impact loading

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2024年第2期32卷 368-382页

作者： Ping Song Jianghai Liu Wenbin Li yiming Li Research Institute of Chemical Defense Academy of Military ScienceBeijing102205China Ministerial Key Laboratory of ZNDY Nanjing University of Science and TechnologyNanjing210094China

In this work,a Cu-10Ta alloy with a copper to tantalum mass ratio of 9:1 is prepared using powder metallurgy *** properties of the alloy,including density,microstructure,melting point,and constant-volume specific heat,are *** the split Hopkinson pressure bar(SHPB)and flyerplate impact experiments,the relationship between equivalent stress and equivalent plastic strain of the material is studied at temperatures of 2.8-82. K and under strain rates of 1×10^(-3)-5.2.10^(3)s^(-1),and the Hugoniot relationship at impact pressures of 1.46-17.2. GPa and impact velocities of 108-942.m/s is *** of the Cu-10Ta microstructure that occurs during high-strain-rate impact is *** indicate that the Cu-10Ta alloy possesses good thermal stability,and at ambient temperatures of up to 50%its melting point,stress softening of less than 15%of the initial strength is observed.A modified J-C constitutive model is employed to accurately predict the variation of yield strength with strain *** strong impact,the copper phase is identified as the primary source of plastic deformation in the Cu-10Ta alloy,while significant deformation of the high-strength tantalum particles is less ***,the longitudinal wave speed D is found to correlate linearly with the particle velocity u upon strong *** reveals that the sound speed and spallation strength of the alloy increase with increasing impact pressure.

关键词： Cu-10Ta SHPB Plastic deformation Flyer impact Hugoniot relationship

Sparse convolutional model with semantic expression for waste electrical appliances recognition

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第9期67卷 2881-2893页

作者： HAN HongGui LIU yiming LI FangYu DU YongPing Faculty of Information Technology Beijing University of TechnologyBeijing 100124China Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing 100124China Engineering Research Center of Digital Community Ministry of EducationBeijing 100124China Beijing Artificial Intelligence Institute and Beijing Laboratory for Intelligent Environmental Protection Beijing 100124China

Deep neural networks play an important role in the recognition of waste electrical appliances. However, deep neural network components still lack reliability in decision-making features. To address this problem, a sparse convolutional model with semantic expression(SCMSE) is proposed. First, a low-rank sparse semantic expression component, combining the benefits of residual networks and sparse representation, is adapted to enhance sparse feature extraction and semantic expression. Second, a reliable network architecture is obtained by iterating the optimal sparse solution, enhancing semantic expression. Finally, the results of visualization experiments on the waste electrical appliances dataset demonstrate that the proposed SCMSE can obtain excellent semantic performance.

关键词： sparse convolutional model deep neural network semantic expression visualization computer vision

Byzantine Robust Federated Learning Scheme Based on Backdoor Triggers

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第5期79卷 2813-2831页

作者： Zheng Yang Ke Gu yiming Zuo School of Computer and Communication Engineering Changsha University of Science and TechnologyChangsha410114China

Federated learning is widely used to solve the problem of data decentralization and can provide privacy protectionfor data owners. However, since multiple participants are required in federated learning, this allows attackers tocompromise. Byzantine attacks pose great threats to federated learning. Byzantine attackers upload maliciouslycreated local models to the server to affect the prediction performance and training speed of the global model. Todefend against Byzantine attacks, we propose a Byzantine robust federated learning scheme based on backdoortriggers. In our scheme, backdoor triggers are embedded into benign data samples, and then malicious localmodels can be identified by the server according to its validation dataset. Furthermore, we calculate the adjustmentfactors of local models according to the parameters of their final layers, which are used to defend against datapoisoning-based Byzantine attacks. To further enhance the robustness of our scheme, each localmodel is weightedand aggregated according to the number of times it is identified as malicious. Relevant experimental data showthat our scheme is effective against Byzantine attacks in both independent identically distributed (IID) and nonindependentidentically distributed (non-IID) scenarios.

关键词： Federated learning Byzantine attacks backdoor triggers

Upper and Lower Bounds of the α-Universal Triple I Method for Unified Interval Implications

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4期79卷 1063-1088页

作者： yiming Tang Jianwei Gao Yifan Huang School of Computer and Information Hefei University of TechnologyHefei230601China Department of Electrical and Computer Engineering University of AlbertaEdmontonAB T6R 2V4Canada

Theα-universal triple I(α-UTI)method is a recognized scheme in the field of fuzzy reasoning,whichwas proposed by our research group *** robustness of fuzzy reasoning determines the quality of reasoning algorithms to a large extent,which is quantified by calculating the disparity between the output of fuzzy reasoning with interference and the output without ***,in this study,the interval robustness(embodied as the interval stability)of theα-UTI method is explored in the interval-valued fuzzy *** begin with,the stability of theα-UTI method is explored for the case of an individual rule,and the upper and lower bounds of its results are estimated,using four kinds of unified interval implications(including the R-interval implication,the S-interval implication,the QL-interval implication and the interval t-norm implication).Through analysis,it is found that theα-UTI method exhibits good interval stability for an individual ***,the stability of theα-UTI method is revealed in the case of multiple rules,and the upper and lower bounds of its outcomes are *** results show that theα-UTI method is stable for multiple rules when four kinds of unified interval implications are used,***,theα-UTI reasoning chain method is presented,which contains a chain structure with multiple *** corresponding solutions and their interval perturbations are *** is found that theα-UTI reasoning chain method is stable in the case of chain *** application examples in affective computing are given to verify the stability of theα-*** summary,through theoretical proof and example verification,it is found that theα-UTImethod has good interval robustness with four kinds of unified interval implications aiming at the situations of an individual rule,multi-rule and reasoning chain.

关键词： Fuzzy reasoning the CRI method the triple I method fuzzy implication interval robustness

SNR-adaptive deep joint source-channel coding scheme for image semantic transmission with convolutional block attention module

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

The Journal of China Universities of Posts and Telecommunications 2024年第1期31卷 1-11页

作者： Yang Yujia Liu yiming Zhang Wenjia Zhang Zhi State Key Laboratory of Networking and Switch Technology Beijing University of Posts and TelecommunicationsBeijing 100876China

With the development of deep learning(DL),joint source-channel coding(JSCC)solutions for end-to-end transmission have gained a lot of *** deep JSCC schemes support dynamically adjusting the rate according to different channel conditions during transmission,enhancing robustness in dynamic wireless ***,most of the existing adaptive JSCC schemes only consider different channel conditions,ignoring the different feature importance in the image processing and *** uniform compression of different features in the image may result in the compromise of critical image details,particularly in low signal-to-noise ratio(SNR)*** address the above issues,in this paper,a dual attention mechanism is introduced and an SNR-adaptive deep JSCC mechanism with a convolutional block attention module(CBAM)is proposed,in which matrix operations are applied to features in spatial and channel dimensions *** proposed solution concatenates the pooling feature with the SNR level and passes it sequentially through the channel attention network and spatial attention network to obtain the importance evaluation *** show that the proposed solution outperforms other baseline schemes in terms of peak SNR(PSNR)and structural similarity(SSIM),particularly in low SNR scenarios or when dealing with complex image content.

关键词： semantic communication joint source-channel coding image transmission

Crust and Upper Mantle Density Structures beneath the Eastern Tianshan Region and Its Tectonic Implications

在线全文

学校读者我要写书评

暂无评论

Journal of Earth Science 2024年第3期35卷 828-838页

作者： yiming Liu Chao Chen Qing Liang Zhengwang Hu Hubei Subsurface Multi-Scale Imaging Key Laboratory School of Geophysics and GeomaticsChina University of GeosciencesWuhan 430074China

The deformation mechanisms of the Tianshan orogenic belt(TOB)are one of the most important unresolved issues in the collision of the Indian and Eurasian *** better understand the lithospheric deformation of the eastern Tianshan orogenic belt,we combined the S-wave tomography and gravity data to develop a three-dimensional(3D)density model of the crust and upper mantle beneath the eastern Tianshan *** show that the crust of the eastern Tianshan is mainly characterized by positive density anomalies,revealing widespread subduction-related magmatism during the *** however have also observed extensive low-density anomalies beneath the eastern Tianshan at depths deeper than~100 km,which is likely linked to a relatively hot *** most fundamental differences of the lithosphere within the eastern Tianshan occur in the uppermost *** uppermost mantle layers in the Bogda Shan and Harlik Shan are relatively *** is likely associated with an eclogite body in the uppermost *** most significant negative anomaly of the uppermost mantle is however found in the Jueluotage tectonic belt and the central Tianshan Block and is possibly associated with depleted mantle *** suggest that these differences related to compositional changes may control the strength of the lithospheric mantle and have affected the uplift of the northern and southern segments of the eastern Tianshan after the Permian.

关键词： constrained inversion gravity anomaly three-dimensional lithospheric density structure eastern Tianshan geodynamics geophysics.

Jet formation and penetration performance of a double-layer charge liner with chemically-deposited tungsten as the inner liner

在线全文

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2024年第3期33卷 374-385页

作者： Bihui Hong Wenbin Li yiming Li Zhiwei Guo Binyou Yan Ministerial Key Laboratory of ZNDY Nanjing University of Science and TechnologyNanjing 210094China State Key Laboratory of Explosion Science and Technology Beijing Institute of TechnologyBeijing 100081China Xiamen Tungsten Co. Ltd.Xiamen 361004China

This paper proposes a type of double-layer charge liner fabricated using chemical vapor deposition(CVD)that has tungsten as its inner *** feasibility of this design was evaluated through penetration ***-layer charge liners were fabricated by using CVD to deposit tungsten layers on the inner surfaces of pure T2.copper *** microstructures of the tungsten layers were analyzed using a scanning electron microscope(SEM).The feasibility analysis was carried out by pulsed X-rays,slug-retrieval test and static penetration *** shaped charge jet forming and penetration law of inner tungsten-coated double-layer liner were studied by numerical simulation *** results showed that the double-layer liners could form well-shaped *** errors between the X-ray test results and the numerical results were within 11.07%.A slug-retrieval test was found that the retrieved slug was similar to a numerically simulated *** with the traditional pure copper shaped charge jet,the penetration depth of the double-layer shaped charge liner increased by 11.4% and>10.8% *** summary,the test results are good,and the numerical simulation is in good agreement with the test,which verified the feasibility of using the CVD method to fabricate double-layer charge liners with a high-density and high-strength refractory metal as the inner liner.

关键词： Shaped charge Chemical vapor deposition Tungsten Double-layer charge liner X-ray Penetration