检索结果-南通市图书馆

Overhead-free Noise-tolerant Federated Learning: A New Baseline

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Machine Intelligence Research 2024年第3期21卷 526-537页

作者： Shiyi Lin Deming Zhai Feilong Zhang Junjun Jiang Xianming Liu Xiangyang Ji department of Computer science and technology Harbin Institute of TechnologyHarbin150000China department of Automation Tsinghua UniversityBeijing100084China

Federated learning (FL) is a promising decentralized machine learning approach that enables multiple distributed clients to train a model jointly while keeping their data private. However, in real-world scenarios, the supervised training data stored in local clients inevitably suffer from imperfect annotations, resulting in subjective, inconsistent and biased labels. These noisy labels can harm the collaborative aggregation process of FL by inducing inconsistent decision boundaries. Unfortunately, few attempts have been made towards noise-tolerant federated learning, with most of them relying on the strategy of transmitting overhead messages to assist noisy labels detection and correction, which increases the communication burden as well as privacy risks. In this paper, we propose a simple yet effective method for noise-tolerant FL based on the well-established co-training framework. Our method leverages the inherent discrepancy in the learning ability of the local and global models in FL, which can be regarded as two complementary views. By iteratively exchanging samples with their high confident predictions, the two models “teach each other” to suppress the influence of noisy labels. The proposed scheme enjoys the benefit of overhead cost-free and can serve as a robust and efficient baseline for noise-tolerant federated learning. Experimental results demonstrate that our method outperforms existing approaches, highlighting the superiority of our method.

关键词： Federated learning noise-label learning privacy-preserving machine learning edge intelligence distributed machine learning

A hybrid memory architecture supporting fine-grained data migration

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 Computer science 2024年第2期18卷 31-41页

作者： Ye CHI Jianhui YUE Xiaofei LIAO Haikun LIU Hai JIN National Engineering Research Center for Big Data technology and System Services Computing Technology and System LabCluster and Grid Computing LabSchool of Computer Science and TechnologyHuazhong University of Science and TechnologyWuhan 430074China department of Computer science Michigan Technological UniversityMichigan 49931USA

Hybrid memory systems composed of dynamic random access memory(DRAM)and Non-volatile memory(NVM)often exploit page migration technologies to fully take the advantages of different memory *** previous proposals usually migrate data at a granularity of 4 KB pages,and thus waste memory bandwidth and DRAM *** this paper,we propose Mocha,a non-hierarchical architecture that organizes DRAM and NVM in a flat address space physically,but manages them in a cache/memory *** the commercial NVM device-Intel Optane DC Persistent Memory Modules(DCPMM)actually access the physical media at a granularity of 256 bytes(an Optane block),we manage the DRAM cache at the 256-byte size to adapt to this feature of *** design not only enables fine-grained data migration and management for the DRAM cache,but also avoids write amplification for Intel Optane *** also create an Indirect Address Cache(IAC)in Hybrid Memory Controller(HMC)and propose a reverse address mapping table in the DRAM to speed up address translation and cache ***,we exploit a utility-based caching mechanism to filter cold blocks in the NVM,and further improve the efficiency of the DRAM *** implement Mocha in an architectural *** results show that Mocha can improve application performance by 8.2%on average(up to 24.6%),reduce 6.9%energy consumption and 25.9%data migration traffic on average,compared with a typical hybrid memory architecture-HSCC.

关键词： non-volatile memory hybrid memory system data migration fine-grained caching

Mechanism in Solidification of a Ternary Nickel Based Alloy

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 Wuhan University of technology(Materials science) 2024年第4期39卷 1018-1024页

作者：田密成博 School of aeronautical Manufacturing Industry Chengdu Aeronautical PolytechnicChengdu 610100China College of Mechanical Engineering North University of ChinaTaiyuan 030051China

The experiment employed the use of melt purification and cyclic superheating technique to achieve maximum undercooling of Ni65Cu31Co4 alloy at ***,high-speed photography techniques were used to capture the process of alloy liquid phase interface migration,and analyzed the relationship between the shape characteristics of the front end of alloy solidification and *** microstructure of the alloy was observed through metallographic microscopy,and the micro-morphological characteristics and evolution of the rapidly solidified microstructure were systematically *** is found that the grain refinement mechanism of Ni-Cu-Co ternary alloy is similar to that of Ni-Cu binary *** refinement at low undercooling is caused by intense dendritic remelting,while grain refinement at high undercooling is attributed to recrystallization,driven by the stress and plastic strain accumulated from the interaction of liquid flow and primary dendrites caused by rapid *** also shows that the addition of the third element Co plays a significant role in solidification rate and re-ignition effect.

关键词： undercooling recalescence Ni-Cu-Co ternary alloy grain refinement recrystallization

维普期刊数据库

Multilevel carbon architecture of subnanoscopic silicon for fast‐charging high‐energy‐density lithium‐ion batteries

在线全文

学校读者我要写书评

暂无评论

Carbon Energy 2024年第4期6卷 256-268页

作者： Meisheng Han Yongbiao Mu Lei Wei Lin Zeng Tianshou Zhao Shenzhen Key Laboratory of Advanced Energy Storage Southern University of Science and TechnologyShenzhenChina SUSTech Energy Institute for Carbon Neutrality Southern University of Science and TechnologyShenzhenChina department of Mechanical and Energy Engineering Southern University of Science and TechnologyShenzhenChina

Silicon(Si)is widely used as a lithium‐ion‐battery anode owing to its high capacity and abundant crustal ***,large volume change upon cycling and poor conductivity of Si cause rapid capacity decay and poor fast‐charging capability limiting its commercial ***,we propose a multilevel carbon architecture with vertical graphene sheets(VGSs)grown on surfaces of subnanoscopically and homogeneously dispersed Si–C composite nanospheres,which are subsequently embedded into a carbon matrix(C/VGSs@Si–C).Subnanoscopic C in the Si–C nanospheres,VGSs,and carbon matrix form a three‐dimensional conductive and robust network,which significantly improves the conductivity and suppresses the volume expansion of Si,thereby boosting charge transport and improving electrode *** VGSs with vast exposed edges considerably increase the contact area with the carbon matrix and supply directional transport channels through the entire material,which boosts charge *** carbon matrix encapsulates VGSs@Si–C to decrease the specific surface area and increase tap density,thus yielding high first Coulombic efficiency and electrode compaction ***,C/VGSs@Si–C delivers excellent Li‐ion storage performances under industrial electrode *** particular,the full cells show high energy densities of 603.5 Wh kg^(−1)and 1685.5 Wh L^(−1)at 0.1 C and maintain 80.7%of the energy density at 3 C.

关键词： fast charging high energy densities lithium‐ion batteries multilevel carbon architecture subnanoscopic silicon anode

Rotational Doppler effect of composite vortex beams with tailored OAM spectra

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Photonics Research 2024年第8期12卷 1665-1672页

作者： YUTIAN LIANG RUIJIAN LI JIE ZHAO XINGYUAN LU TONG LIU ZHENGLIANG LIU YUAN REN CHENGLIANG ZHAO department of Aerospace science and technology Space Engineering UniversityBeijing 101416China School of Physical science and technology Soochow UniversitySuzhou 215006China department of Basic Course Space Engineering UniversityBeijing 101416China

There recently has been increasing interest in the research and application of the rotational Doppler effect(RDE),which paves a promising way to detect rotating objects *** order to obtain more information about the rotating object from the rotational Doppler signal,composite vortex beams by coaxial superposition of orbital angular momentum(OAM)modes are often used as the probe ***,to the best of our knowledge,the RDE of composite vortex beams with arbitrary OAM spectra has not yet been comprehensively *** this paper,the correspondence between the OAM spectrum of a probe beam and the frequency spectrum of a rotational Doppler signal is theoretically *** is explicitly revealed that the RDE frequency spectrum of scattered light is related to the product of two autocorrelation functions:one from the OAM spectrum of probe beam and the other from the spiral spectrum of rotating *** the basis of this relation,one can regulate the RDE frequency spectrum on demand via tailoring the OAM spectrum of the probe *** a proof of concept we design a special composite vortex beam to eliminate the broadening of the RDE spectrum induced by *** findings are of practical value in applications such as remote sensing and optical metrology.

关键词： OAM composite rotating

维普期刊数据库

Drag reduction characteristics of recirculation flow at rocket base in an RBCC engine under ramjet/scramjet mode

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第3期37卷 104-115页

作者： Menglei LI Mingbo SUN Peibo LI Daoning YANG Bin AN Yizhi YAO Jikai CHEN Taiyu WANG Jiaoru WANG Hypersonic technology Laboratory National University of Defense TechnologyChangsha 410073China department of Aerospace science and technology Space Engineering UniversityBeijing 101416China

To reduce the drag generated by the recirculation flow at the rocket base in a RocketBased Combined Cycle(RBCC)engine operating in the ramjet/scramjet mode,a novel annular rocket RBCC engine based on a central plug cone was *** performance loss mechanism caused by the recirculation flow at the rocket base and the influence of the plug cone configuration on the thrust performance were *** indicated that the recirculation flow at the rocket base extended through the entire combustor,which creates an extensive range of the"low-kineticenergy zone"at the center and leads to an engine thrust *** plug cone serving as a surface structure had a restrictive effect on the internal flow of the engine,making it smoothly transit at the position of the large separation *** model RBCC engine could achieve a maximum thrust augmentation of 37.6%with a long plug cone that was twice diameter of the inner ***,a shorter plug cone that was half diameter of the inner isolator proved less effective at reducing the recirculation flow for a supersonic flow and induced an undesirable flow fraction that diminished the thrust ***,the effectiveness of the plug cone increased with the flight Mach number,indicating that it could further broaden the operating speed range of the scramjet mode.

关键词： Rocket-based combined cycle Ramjet/scramjet mode Plug cone Drag reduction k-x SST turbulent model

维普期刊数据库

Highly anti-corrosive NiFe LDHs–NiFe alloy hybrid enables long-term stable alkaline seawater electrolysis

在线全文

学校读者我要写书评

暂无评论

Rare Metals 2024年第9期43卷 4321-4332页

作者： Jia-Hong Li Hao Chen Si-Hang You Gan-Xin Yang Peng Liu Meng-Qi Gao Shu-Guang Chen Fei-Fei Zhang School of Materials science and Engineering Changsha University of Science and Technology department of Materials science and Engineering Faculty of Engineering National University of Singapore School of Materials science and Engineering Jilin University

Owing to the significant potential of alkalin seawater electrolysis for converting surplus power into eco friendly hydrogen fuel,we developed bifunctional elec trodes that integrate low-crystalline NiFe LDHs and amorphous NiFe alloy on a Ni foam （NF） substrate to enhance this *** by the battery-like charac teristics of NiFe LDHs,an anti-corrosive and active oute layer of NiFevacOOH continuously forms over time in th hybrid on the anode for the oxygen evolution reaction（OER）,effectively mitigating powder shedding caused by corrosion induced by multiple anions in *** while,the strong bond between the hybrid and the NF substrate maintains intact hybrid coatings to ensure a rel atively high overall conductivity of the electrodes,signif icantly reducing the negative effects of structura degradation during the OER and hydrogen evolution reaction （HER）,as well as the accumulation of contami nants on the electrode *** long-term tests,thes bifunctionalhybridelectrodesmaintained stable performance,even at a high current density o500 mA·*** cell voltage increased by only 88 m V over 1000 h to 1.970 V during saline electrolysis and by103 mV over 500 h to 2.062 V during seawater electroly ***,this study provides valuable insights into efficient and stable seawater electrolysis using NiFe LDHs–NiFe alloy hybrids.

关键词： Bifunctional Corrosion NiFe LDHs Seawater electrolysis Stability

Manipulating photogenerated electron flow in nickel single‐atom catalysts for photocatalytic CO_(2)reduction into tunable syngas

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Carbon Energy 2024年第8期6卷 205-213页

作者： Yida Zhang Qingyu Wang Lihui Wu Haibin Pan Chengyuan Liu Yue Lin Gongming Wang Xusheng Zheng National Synchrotron Radiation Laboratory University of Science and Technology of ChinaHefeiAnhuiChina College of Chemistry and Materials science University of Science and Technology of ChinaHefeiAnhuiChina department of Chemistry Hefei National Research Center for Physical Sciences at the MicroscaleUniversity of Science and Technology of ChinaHefeiAnhuiChina

The key to designing photocatalysts is to orient the migration of photogenerated electrons to the target active sites rather than dissipate at inert ***,we demonstrate that the doping of phosphorus(P)significantly enriches photogenerated electrons at Ni active sites and enhances the performance for CO_(2)reduction into *** photocatalytic CO_(2)reduction,Ni single‐atom‐anchored P‐modulated carbon nitride showed an impressive syngas yield rate of 85μmol gcat^(−1)h^(−1)and continuously adjustable CO/H_(2)ratios ranging from 5:1 to 1:2,which exceeded those of most of the reported carbon nitride‐based single‐atom *** studies reveal that P doping improves the conductivity of catalysts,which promotes photogenerated electron transfer to the Ni active sites rather than dissipate randomly at low‐activity nonmetallic sites,facilitating the CO_(2)‐to‐syngas photoreduction process.

关键词： carbon nitride CO_(2)photoreduction electron flow Ni single atoms syngas

AutoRhythmAI: A Hybrid Machine and Deep Learning Approach for Automated Diagnosis of Arrhythmias

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2期78卷 2137-2158页

作者： S.Jayanthi S.Prasanna Devi department of Computer science and Engineering SRM Institute of Science and Technology-Vadapalani CampusChennaiTamil NaduIndia

In healthcare,the persistent challenge of arrhythmias,a leading cause of global mortality,has sparked extensive research into the automation of detection using machine learning(ML)***,traditional ML and AutoML approaches have revealed their limitations,notably regarding feature generalization and automation *** glaring research gap has motivated the development of AutoRhythmAI,an innovative solution that integrates both machine and deep learning to revolutionize the diagnosis of *** approach encompasses two distinct pipelines tailored for binary-class and multi-class arrhythmia detection,effectively bridging the gap between data preprocessing and model *** validate our system,we have rigorously tested AutoRhythmAI using a multimodal dataset,surpassing the accuracy achieved using a single dataset and underscoring the robustness of our *** the first pipeline,we employ signal filtering and ML algorithms for preprocessing,followed by data balancing and split for *** second pipeline is dedicated to feature extraction and classification,utilizing deep learning ***,we introduce the‘RRI-convoluted trans-former model’as a novel addition for binary-class *** ensemble-based approach then amalgamates all models,considering their respective weights,resulting in an optimal model *** our study,the VGGRes Model achieved impressive results in multi-class arrhythmia detection,with an accuracy of 97.39%and firm performance in precision(82.13%),recall(31.91%),and F1-score(82.61%).In the binary-class task,the proposed model achieved an outstanding accuracy of 96.60%.These results highlight the effectiveness of our approach in improving arrhythmia detection,with notably high accuracy and well-balanced performance metrics.

关键词： Automated machine learning neural networks deep learning arrhythmias